attribute

attribute_filter_type

condition_class

include_special_attributes

invert_selection

measure_types

method

parameter_string

sample

sample_probability

sample_ratio

sample_size

svm_type

target_role

containsColumn(?Din,?Column) -> AttributeManipulation(?this)

DataTable(?Din) -> new(?this), DataTableProcessing(?this), uses(?this, ?Din)

uses(?this,?D), Clusterer(?this) -> copy(?Dout,?D, {}), newFor(?ClusterColumn, ?Dout), CategorialColumn(?ClusterColumn), columnHasType(?ClusterColumn,iCategorial), amountOfMissingValues(?ClusterColumn, 0), newFor(?ClusterAttribute, ?ClusterColumn), Attribute(?ClusterAttribute), representsAttribute(?ClusterColumn, ?ClusterAttribute), containsColumn(?Dout,?ClusterColumn), clusterID(?Dout, ?ClusterAttribute), produces(?this, ?Dout)

OutlierDetection(?this), uses(?this, ?Din) -> copy(?Dout,?Din, {}), newFor(?OutlierColumn, ?Dout), DataColumn(?OutlierColumn), amountOfMissingValues(?OutlierColumn, 0), newFor(?OutlierAttribute, ?OutlierColumn), Attribute(?OutlierAttribute), attributeName(?OutlierAttribute, "outlier"), representsAttribute(?OutlierColumn, ?OutlierAttribute), containsColumn(?Dout,?OutlierColumn), outlier(?Dout, ?OutlierAttribute), produces(?this, ?Dout)

1 1

amountOfRows(?Din, ?NumRows) -> RM_AbsoluteSampling_1000(?this), simpleParameter_sample_size(?this, "1000")

RM_AbsoluteSampling_1000(?this), uses(?this, ?Din), simpleParameter_sample_size(?this, "1000") -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), amountOfRows(?Dout, 1000), produces(?this, ?Dout)

Absolute sampling operator. This operator takes a random sample with the given size. For example, if the sample size is set to 50, the result will have exactly 50 examples randomly drawn from the complete data set. Please note that this operator does not sample during a data scan but jumps to the rows. It should therefore only be used in case of memory data management and not, for example, for database or file management. Parameters: hasParameterKey sample_size hasDefalutValue 100 comment The number of examples which should be sampled hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

absolutesampling

1 1

amountOfRows(?Din, ?NumRows) -> RM_AbsoluteStratifiedSampling(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Stratified sampling operator. This operator performs a random sampling of a given size. In contrast to the simple sampling operator, this operator performs a stratified sampling for data sets with nominal label attributes, i.e. the class distributions remains (almost) the same after sampling. Hence, this operator cannot be applied on data sets without a label or with a numerical label. In these cases a simple sampling without stratification is performed. In some cases it might happen that not the exact desired number of examples is sampled, e.g. if the desired number is 100 from three qually distributed classes the resulting number will be 99 (33 of each class). Parameters: hasParameterKey sample_size hasDefalutValue 100 comment The number of examples which should be sampled hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

RM_AbsoluteStratifiedSampling(?this), uses(?this, ?Din) -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), produces(?this, ?Dout)

absolutestratifiedsampling

1 1

attributebasedvote

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_AttributeBasedVote(?this)

AttributeBasedVotingLearner is very lazy. Actually it does not learn at all but creates an AttributeBasedVotingModel. This model simply calculates the average of the attributes as prediction (for regression) or the mode of all attribute values (for classification). AttributeBasedVotingLearner is especially useful if it is used on an example set created by a meta learning scheme, e.g. by Vote. Parameters:

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_AttributeBasedVote(?this) ->

1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_BasicRuleLearner(?this)

This operator builds an unpruned rule set of classification rules. It is based on the paper Cendrowska, 1987: PRISM: An algorithm for inducing modular rules.

RM_BasicRuleLearner(?this) ->

Parameters: hasParameterKey pureness hasDefalutValue 0.9 comment The desired pureness, i.e. the necessary amount of the major class in a covered subset in order become pure.

basicrulelearner

1 1

RM_CHAID(?this) ->

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_CHAID(?this)

The CHAID decision tree learner works like the DecisionTreeLearner with one exception: it used a chi squared based criterion instead of the information gain or gain ratio criteria. Parameters: hasParameterKey minimal_size_for_split hasDefalutValue 4 comment The minimal size of a node in order to allow a split. hasParameterKey minimal_leaf_size hasDefalutValue 2 comment The minimal size of all leaves. hasParameterKey minimal_gain hasDefalutValue 0.1 comment The minimal gain which must be achieved in order to produce a split. hasParameterKey maximal_depth hasDefalutValue 20 comment The maximum tree depth (-1: no bound) hasParameterKey confidence hasDefalutValue 0.25 comment The confidence level used for the pessimistic error calculation of pruning. hasParameterKey number_of_prepruning_alternatives hasDefalutValue 3 comment The number of alternative nodes tried when prepruning would prevent a split. hasParameterKey no_pre_pruning hasDefalutValue false comment Disables the pre pruning and delivers a tree without any prepruning. hasParameterKey no_pruning hasDefalutValue false comment Disables the pruning and delivers an unpruned tree.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

chaid

1 1 1

RM_DBSCAN(?this) ->

[DataTable and (inputColumn only (NormalizedScalarColumn or CategorialColumn))](?D) -> RM_DBSCAN(?this)

This operator provides the DBScan cluster algorithm. If no id attribute is present, the operator will create one.

dbscan

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey epsilon hasDefalutValue 1.0 comment Specifies the size of neighbourhood. hasParameterKey min_points hasDefalutValue 5 comment The minimal number of points forming a cluster. hasParameterKey add_cluster_attribute hasDefalutValue true comment Indicates if a cluster id is generated as new special attribute. hasParameterKey add_as_label hasDefalutValue false comment Should the cluster values be added as label. hasParameterKey remove_unlabeled hasDefalutValue false comment Delete the unlabeled examples. hasParameterKey measure_types hasDefalutValue MixedMeasures comment The measure type hasParameterKey mixed_measure hasDefalutValue MixedEuclideanDistance comment Select measure hasParameterKey nominal_measure hasDefalutValue NominalDistance comment Select measure hasParameterKey numerical_measure hasDefalutValue EuclideanDistance comment Select measure hasParameterKey divergence hasDefalutValue GeneralizedIDivergence comment Select divergence hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b. 1 1

decision_stump

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator learns decision stumps, i.e. a small decision tree with only one single split. This decision stump works on both numerical and nominal attributes. Parameters: hasParameterKey criterion hasDefalutValue gain_ratio comment Specifies the used criterion for selecting attributes and numerical splits. hasParameterKey minimal_size_for_split hasDefalutValue 4 comment The minimal size of a node in order to allow a split. hasParameterKey minimal_leaf_size hasDefalutValue 1 comment The minimal size of all leaves. hasParameterKey minimal_gain hasDefalutValue 0.1 comment The minimal gain which must be achieved in order to produce a split.

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_Decision_Stump(?this)

RM_Decision_Stump(?this) ->

1 1

decision_tree

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator learns decision trees from both nominal and numerical data. Decision trees are powerful classification methods which often can also easily be understood. In order to classify an example, the tree is traversed bottom-down. Every node in a decision tree is labelled with an attribute. The example's value for this attribute determines which of the outcoming edges is taken. For nominal attributes, we have one outgoing edge per possible attribute value, and for numerical attribtues the outgoing edges are labelled with disjoint ranges. This decision tree learner works similar to Quinlan's C4.5 or CART. Roughly speaking, the tree induction algorithm works as follows. Whenever a new node is created at a certain stage, an attribute is picked to maximise the discriminative power of that node with respect to the examples assigned to the particular subtree. This discriminative power is measured by a criterion which can be selected by the user (information gain, gain ratio, gini index, etc.). The algorithm stops in various cases: <ul> <li> No attribute reaches a certain threshold (minimum_gain). </li> <li> The maximal depth is reached. </li> <li> There are less than a certain number of examples (minimal_size_for_split) in the current subtree. </li> </ul> Finally, the tree is pruned, i.e. leaves that do not add to the discriminative power of the whole tree are removed.

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_Decision_Tree(?this)

RM_Decision_Tree(?this) ->

1 1

default_model

Parameters: hasParameterKey method hasDefalutValue median comment The method to compute the default. hasParameterKey constant hasDefalutValue 0.0 comment Value returned when method = constant. hasParameterKey attribute_name hasDefalutValue comment The attribute to get the predicted value from.

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_Default_Model(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Default_Model(?this) ->

This learner creates a model, that will simply predict a default value for all examples, i.e. the average or median of the true labels (or the mode in case of classification) or a fixed specified value. This learner can be used to compare the results of "real" learning schemes with guessing. 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

detect_outlier_cof

This operator performs a Class Outlier Factor (COF) search. COF outliers or Class Outliers method search for observations (objects) those that arouse suspicions, taking into account the class labels according to the definition of Class Outlier by Hewaihi and Saad in "A comparative Study of Outlier Mining and Class Outlier Mining", CS Letters, Vol 1, No 1 (2009)", and "Class Outliers Mining: Distance-Based Approach", International Journal of Intelligent Systems and Technologies, Vol. 2, No. 1, pp 55-68, 2007". It detects rare / exceptional / suspicious cases with respect to a group of similar cases. The main concept of ECODB (Enhanced Class Outlier - Distance Based) algorithm is to rank each instance in the dataset D given the parameters N (top N class outliers), and K (the number of nearest neighbors. The Rank finds out the rank of each instance using the formula (COF = PCL(T,K) - norm(deviation(T)) + norm(kDist(T))). where PCL(T,K) is the Probability of the class label of the instance T with respect to the class labels of its K Nearest Neighbors. and norm(Deviation(T)) and norm(KDist(T)) are the normalized value of Deviation(T) and KDist(T) respectively and their value fall into the range [0 - 1]. Deviation(T) is how much the instance T deviates from instances of the same class, and computed by summing the distances between the instance T and every instance belong to the same class of the instance. KDist(T) is the summation of distance between the instance T and its K nearest neighbors. The ECODB algorithm maintains a list of only the instances of the top N class outliers. The less is the value of COF of an instance, the higher is the priority of the instance to be a class outlier. The operator supports mixed euclidian distance. The Operator takes an example set and passes it on with an boolean top-n COF outlier status in a new boolean-valued special outlier attribute indicating true (outlier) and false (no outlier), and another special attribute "COF Factor" which measures the degree of being Class Outlier for an object. Parameters: hasParameterKey number_of_neighbors hasDefalutValue 7 comment Specifies the k value for the k-th nearest neighbours to be the analyzed. (default value is 10, minimum 1 and max is set to 1 million) hasParameterKey number_of_class_outliers hasDefalutValue 10 comment The number of top-n Class Outliers to be looked for.(default value is 10, minimum 2 (internal reasons) and max is set to 1 million) hasParameterKey measure_types hasDefalutValue MixedMeasures comment The measure type hasParameterKey mixed_measure hasDefalutValue MixedEuclideanDistance comment Select measure hasParameterKey nominal_measure hasDefalutValue NominalDistance comment Select measure hasParameterKey numerical_measure hasDefalutValue EuclideanDistance comment Select measure hasParameterKey divergence hasDefalutValue GeneralizedIDivergence comment Select divergence hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b.

RM_Detect_Outlier_COF(?this) -> BinaryColumn(?OutlierColumn), columnHasType(?OutlierColumn, iBinary), possibleValue(?OutlierColumn, "true"), possibleValue(?OutlierColumn, "false")

-> RM_Detect_Outlier_COF(?this)

1 1 Parameters: hasParameterKey distance hasDefalutValue comment The distance for objects. hasParameterKey proportion hasDefalutValue comment The proportion of objects related to D. hasParameterKey distance_function hasDefalutValue euclidian distance comment Indicates which distance function will be used for calculating the distance between two objects

-> RM_Detect_Outlier_Densities(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

detect_outlier_densities

This operator is a DB outlier detection algorithm which calculates the DB(p,D)-outliers for an ExampleSet passed to the operator. DB(p,D)-outliers are Distance based outliers according to Knorr and Ng. A DB(p,D)-outlier is an object to which at least a proportion of p of all objects are farer away than distance D. It implements a global homogenous outlier search. Currently, the operator supports cosine, sine or squared distances in addition to the usual euclidian distance which can be specified by the corresponding parameter. The operator takes two other real-valued parameters p and D. Depending on these parameters, search objects will be created from the examples in the ExampleSet passed to the operator. These search objects will be added to a search space which will perform the outlier search according to the DB(p,D) scheme. The Outlier status (boolean in its nature) is written to a new special attribute "Outlier" and is passed on with the example set.

RM_Detect_Outlier_Densities(?this) -> BinaryColumn(?OutlierColumn), columnHasType(?OutlierColumn, iBinary), possibleValue(?OutlierColumn, "true"), possibleValue(?OutlierColumn, "false")

1 1 Parameters: hasParameterKey number_of_neighbors hasDefalutValue 10 comment Specifies the k value for the k-th nearest neighbours to be the analyzed.(default value is 10, minimum 1 and max is set to 1 million) hasParameterKey number_of_outliers hasDefalutValue 10 comment The number of top-n Outliers to be looked for.(default value is 10, minimum 2 (internal reasons) and max is set to 1 million) hasParameterKey distance_function hasDefalutValue euclidian distance comment choose which distance function will be used for calculating the distance between two objects This operator performs a D^k_n Outlier Search according to the outlier detection approach recommended by Ramaswamy, Rastogi and Shim in "Efficient Algorithms for Mining Outliers from Large Data Sets". It is primarily a statistical outlier search based on a distance measure similar to the DB(p,D)-Outlier Search from Knorr and Ng. But it utilizes a distance search through the k-th nearest neighbourhood, so it implements some sort of locality as well. The method states, that those objects with the largest distance to their k-th nearest neighbours are likely to be outliers respective to the data set, because it can be assumed, that those objects have a more sparse neighbourhood than the average objects. As this effectively provides a simple ranking over all the objects in the data set according to the distance to their k-th nearest neighbours, the user can specify a number of n objects to be the top-n outliers in the data set. The operator supports cosine, sine or squared distances in addition to the euclidian distance which can be specified by a distance parameter. The Operator takes an example set and passes it on with an boolean top-n D^k outlier status in a new boolean-valued special outlier attribute indicating true (outlier) and false (no outlier).

-> RM_Detect_Outlier_Distances(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

detect_outlier_distances

RM_Detect_Outlier_Distances(?this) -> BinaryColumn(?OutlierColumn), columnHasType(?OutlierColumn, iBinary), possibleValue(?OutlierColumn, "true"), possibleValue(?OutlierColumn, "false")

1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator performs a LOF outlier search. LOF outliers or outliers with a local outlier factor per object are density based outliers according to Breuning, Kriegel, et al. The approach to find those outliers is based on measuring the density of objects and its relation to each other (referred to as local reachability density). Based on the average ratio of the local reachability density of an object and its k-nearest neighbours (e.g. the objects in its k-distance neighbourhood), a local outlier factor (LOF) is computed. The approach takes a parameter MinPts (actually specifying the "k") and it uses the maximum LOFs for objects in a MinPts range (lower bound and upper bound to MinPts). Currently, the operator supports cosine, sine or squared distances in addition to the usual euclidian distance which can be specified by the corresponding parameter. In the first step, the objects are grouped into containers. For each object, using a radius screening of all other objects, all the available distances between that object and another object (or group of objects) on the (same) radius given by the distance are associated with a container. That container than has the distance information as well as the list of objects within that distance (usually only a few) and the information, how many objects are in the container. In the second step, three things are done: (1) The containers for each object are counted in acending order according to the cardinality of the object list within the container (= that distance) to find the k-distances for each object and the objects in that k-distance (all objects in all the subsequent containers with a smaller distance). (2) Using this information, the local reachability densities are computed by using the maximum of the actual distance and the k-distance for each object pair (object and objects in k-distance) and averaging it by the cardinality of the k-neighbourhood and than taking the reciprocal value. (3) The LOF is computed for each MinPts value in the range (actually for all up to upper bound) by averaging the ratio between the MinPts-local reachability-density of all objects in the k-neighbourhood and the object itself. The maximum LOF in the MinPts range is passed as final LOF to each object. Afterwards LOFs are added as values for a special real-valued outlier attribute in the example set which the operator will return. Parameters: hasParameterKey minimal_points_lower_bound hasDefalutValue 10 comment The lower bound for MinPts for the Outlier test (default value is 10) hasParameterKey minimal_points_upper_bound hasDefalutValue 20 comment The upper bound for MinPts for the Outlier test (default value is 20) hasParameterKey distance_function hasDefalutValue euclidian distance comment choose which distance function will be used for calculating the distance between two objects

-> RM_Detect_Outlier_LOF(?this)

detect_outlier_lof

RM_Detect_Outlier_LOF(?this) -> NormalizedScalarColumn(?OutlierColumn), columnHasType(?OutlierColumn, iScalar), minimumValue(?OutlierColumn, 0), maximumValue(?OutlierColumn, 1)

1 1 1

RM_Discretize_by_Binning_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Binning_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

discretize_by_bins

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 1 This operator discretizes all numeric attributes in the dataset into nominal attributes. This discretization is performed by simple binning, i.e. the specified number of equally sized bins is created and the numerical values are simply sorted into those bins. Skips all special attributes including the label.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Discretize_by_Binning_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

discretize_by_bins

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Binning_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

discretize_by_entropy

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Entropy_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

RM_Discretize_by_Entropy_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

This operator discretizes all numeric attributes in the dataset into nominal attributes. The discretization is performed by selecting a bin boundary minimizing the entropy in the induced partitions. The method is then applied recursively for both new partitions until the stopping criterion is reached. For Details see a) Multi-interval discretization of continued-values attributes for classification learning (Fayyad,Irani) and b) Supervised and Unsupervised Discretization (Dougherty,Kohavi,Sahami). Skips all special attributes including the label. Please note that this operator automatically removes all attributes with only one range (i.e. those attributes which are not actually discretized since the entropy criterion is not fulfilled). This behavior can be controlled by the remove_useless parameter. 1 1 1 Parameters: hasParameterKey return_preprocessing_model hasDefalutValue false comment Indicates if the preprocessing model should also be returned hasParameterKey create_view hasDefalutValue false comment Create View to apply preprocessing instead of changing the data hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey remove_useless hasDefalutValue true comment Indicates if useless attributes, i.e. those containing only one single range, should be removed. hasParameterKey range_name_type hasDefalutValue long comment Indicates if long range names including the limits should be used. hasParameterKey automatic_number_of_digits hasDefalutValue true comment Indicates if the number of digits should be automatically determined for the range names. hasParameterKey number_of_digits hasDefalutValue -1 comment The minimum number of digits used for the interval names (-1: determine minimal number automatically).

discretize_by_entropy

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Entropy_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Discretize_by_Entropy_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

1 1 1

RM_Discretize_by_Frequency_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

discretize_by_frequency

This operator discretizes all numeric attributes in the dataset into nominal attributes. This discretization is performed by equal frequency binning, i.e. the thresholds of all bins is selected in a way that all bins contain the same number of numerical values. The number of bins is specified by a parameter, or, alternatively, is calculated as the square root of the number of examples with non-missing values (calculated for every single attribute). Skips all special attributes including the label. Note that it is possible to get bins with different numbers of examples. This might occur, if the attributes's values are not unique, since the algorithm can not split between examples with same value.

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Frequency_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Frequency_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

RM_Discretize_by_Frequency_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

discretize_by_frequency

1 1 1

discretize_by_size

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Size_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Discretize_by_Size_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

This operator discretizes all numeric attributes in the dataset into nominal attributes. This discretization is performed by binning examples into bins of same size. The specified number of equally sized bins is created and the numerical values are simply sorted into those bins, so that all bins contain the same number of examples. Skips all special attributes including the label. 1 1 1

discretize_by_size

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_Size_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Discretize_by_Size_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_User_Specification_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

RM_Discretize_by_User_Specification_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

This operator discretizes a numerical attribute to either a nominal or an ordinal attribute. The numerical values are mapped to the classes according to the thresholds specified by the user. The user can define the classes by specifying the upper limits of each class. The lower limit of the next class is automatically specified as the upper limit of the previous one. A parameter defines to which adjacent class values that are equal to the given limits should be mapped. If the upper limit in the last list entry is not equal to Infinity, an additional class which is automatically named is added. If a '?' is given as class value the according numerical values are mapped to unknown values in the resulting attribute.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

discretize_by_user_specification

Parameters: hasParameterKey return_preprocessing_model hasDefalutValue false comment Indicates if the preprocessing model should also be returned hasParameterKey create_view hasDefalutValue false comment Create View to apply preprocessing instead of changing the data hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey attribute_type hasDefalutValue nominal comment Attribute type of the discretized attribute. hasParameterKey classes hasDefalutValue first␝-Infinity␞last␝Infinity comment Defines the classes and the upper limits of each class. 1 1 1 Parameters: hasParameterKey return_preprocessing_model hasDefalutValue false comment Indicates if the preprocessing model should also be returned hasParameterKey create_view hasDefalutValue false comment Create View to apply preprocessing instead of changing the data hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey attribute_type hasDefalutValue nominal comment Attribute type of the discretized attribute. hasParameterKey classes hasDefalutValue first␝-Infinity␞last␝Infinity comment Defines the classes and the upper limits of each class. This operator discretizes a numerical attribute to either a nominal or an ordinal attribute. The numerical values are mapped to the classes according to the thresholds specified by the user. The user can define the classes by specifying the upper limits of each class. The lower limit of the next class is automatically specified as the upper limit of the previous one. A parameter defines to which adjacent class values that are equal to the given limits should be mapped. If the upper limit in the last list entry is not equal to Infinity, an additional class which is automatically named is added. If a '?' is given as class value the according numerical values are mapped to unknown values in the resulting attribute.

RM_Discretize_by_User_Specification_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

discretize_by_user_specification

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Discretize_by_User_Specification_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

fast_large_margin

Applies a fast margin learner based on the linear support vector learning scheme proposed by R.-E. Fan, K.-W. Chang, C.-J. Hsieh, X.-R. Wang, and C.-J. Lin. Although the result is similar to those delivered by classical SVM or logistic regression implementations, this linear classifier is able to work on data set with millions of examples and attributes.

RM_Fast_Large_Margin(?this) ->

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Fast_Large_Margin(?this)

Parameters: hasParameterKey solver hasDefalutValue L2 SVM Dual comment The solver type for this fast margin method. hasParameterKey C hasDefalutValue 1.0 comment The cost parameter C for c_svc, epsilon_svr, and nu_svr. hasParameterKey epsilon hasDefalutValue 0.01 comment Tolerance of termination criterion. hasParameterKey class_weights hasDefalutValue comment The weights w for all classes, i.e. set the parameters C of each class w * C (empty: using 1 for all classes where the weight was not defined). hasParameterKey use_bias hasDefalutValue true comment Indicates if an intercept value should be calculated. 1 1 Parameters: hasParameterKey condition_class hasDefalutValue all comment Implementation of the condition. hasParameterKey parameter_string hasDefalutValue comment Parameter string for the condition, e.g. 'attribute=value' for the AttributeValueFilter. hasParameterKey invert_filter hasDefalutValue false comment Indicates if only examples should be accepted which would normally filtered.

RM_Filter_Examples_attribute_value_filter_outlier_0_3(?this), uses(?this, ?Din), outlier(?Din, ?OutlierColumn), representsAttribute(?OutlierColumn,?OutlierAttribute), simpleParameter_condition_class(?this, "attribute_value_filter"), simpleParameter_parameter_string(?this, "outlier_0_3") -> copy(?Dout, ?Din, {containsColumn(?Din, ?OutlierColumn), outlier(?Din, ?OutlierAttribute), amountOfRows(?Din, ?_)}), copy(?NewOutlierColumn, ?OutlierColumn, {maximumValue(?OutlierColumn, ?_)}), representsAttribute(?NewOutlierColumn, ?OutlierAttribute), maximumValue(?NewOutlierColumn, 0.3), produces(?this, ?Dout)

outlier(?Din, ?OutlierAttribute), containsColumn(?Din, ?OutlierColumn), representsAttribute(?OutlierColumn, ?OutlierAttribute) -> RM_Filter_Examples_attribute_value_filter_outlier_0_3(?this), simpleParameter_condition_class(?this, "attribute_value_filter"), simpleParameter_parameter_string(?this, "outlier_0_3")

filter_examples

This operator takes an ExampleSet as input and returns a new ExampleSet including only the Examples that fulfill a condition. By specifying an implementation of Condition and a parameter string, arbitrary filters can be applied. Users can implement their own conditions by writing a subclass of the above class and implementing a two argument constructor taking an ExampleSet and a parameter string. This parameter string is specified by the parameter <code>parameter_string</code>. Instead of using one of the predefined conditions users can define their own implementation with the fully qualified class name. For "attribute_value_condition" the parameter string must have the form <code>attribute op value</code>, where attribute is a name of an attribute, value is a value the attribute can take and op is one of the binary logical operators similar to the ones known from Java, e.g. greater than or equals. Please note your can define a logical OR of several conditions with || and a logical AND of two conditions with two ampers and - or simply by applying several ExampleFilter operators in a row. Please note also that for nominal attributes you can define a regular expression for value of the possible equal and not equal checks. For "unknown_attributes" the parameter string must be empty. This filter removes all examples containing attributes that have missing or illegal values. For "unknown_label" the parameter string must also be empty. This filter removes all examples with an unknown label value.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 Parameters: hasParameterKey condition_class hasDefalutValue all comment Implementation of the condition. hasParameterKey parameter_string hasDefalutValue comment Parameter string for the condition, e.g. 'attribute=value' for the AttributeValueFilter. hasParameterKey invert_filter hasDefalutValue false comment Indicates if only examples should be accepted which would normally filtered.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

filter_examples

RM_Filter_Examples_no_missing_attributes(?this), uses(?this, ?Din), containsColumn(?Din, ?Col), representsAttribute(?Col, ?Att), simpleParameter_condition_class(?this, "no_missing_attributes") -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_), containsColumn(?Din, ?_), hasAttribute(?Din, ?_)}), copy(?NewCol, ?Col, {amountOfMissingValues(?Col, ?_), representsAttribute(?Col,?_)}), copy(?NewAtt, ?Att, {}), containsColumn(?Dout, ?NewCol), hasAttribute(?Dout, ?NewAtt), representsAttribute(?NewCol, ?NewAtt), amountOfMissingValues(?NewCol, 0), produces(?this, ?Dout)

containsColumn(?Din, ?Col), representsAttribute(?Col, ?Att), not({MissingValueFreeColumn(?Col)}) -> RM_Filter_Examples_no_missing_attributes(?this), simpleParameter_condition_class(?this, "no_missing_attributes")

1 1

gaussian_process

Gaussian Process (GP) Learner. The GP is a probabilistic method both for classification and regression.

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Gaussian_Process(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey kernel_type hasDefalutValue rbf comment The kind of kernel. hasParameterKey kernel_lengthscale hasDefalutValue 3.0 comment The lengthscale used in all kernels. hasParameterKey kernel_degree hasDefalutValue 2.0 comment The degree used in the poly kernel. hasParameterKey kernel_bias hasDefalutValue 1.0 comment The bias used in the poly kernel. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The SVM kernel parameter sigma1 (Epanechnikov, Gaussian Combination, Multiquadric). hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The SVM kernel parameter sigma2 (Gaussian Combination). hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The SVM kernel parameter sigma3 (Gaussian Combination). hasParameterKey kernel_shift hasDefalutValue 1.0 comment The SVM kernel parameter shift (Multiquadric). hasParameterKey kernel_a hasDefalutValue 1.0 comment The SVM kernel parameter a (neural). hasParameterKey kernel_b hasDefalutValue 0.0 comment The SVM kernel parameter b (neural). hasParameterKey max_basis_vectors hasDefalutValue 100 comment Maximum number of basis vectors to be used. hasParameterKey epsilon_tol hasDefalutValue 1.0E-7 comment Tolerance for gamma induced projections hasParameterKey geometrical_tol hasDefalutValue 1.0E-7 comment Tolerance for geometry induced projections

RM_Gaussian_Process(?this) ->

1 1

hyper_hyper

RM_Hyper_Hyper(?this) ->

Parameters: hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Hyper_Hyper(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This is a minimal SVM implementation. The model is built with only one positive and one negative example. Typically this operator is used in combination with a boosting method. 1 1 This operator learns decision trees without pruning using nominal attributes only. Decision trees are powerful classification methods which often can also easily be understood. This decision tree learner works similar to Quinlan's ID3.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_ID3(?this) ->

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_ID3(?this)

id3

1 1

id3numerical

This operator learns decision trees without pruning using both nominal and numerical attributes. Decision trees are powerful classification methods which often can also easily be understood. This decision tree learner works similar to Quinlan's ID3.

RM_ID3Numerical(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_ID3Numerical(?this)

RM_IteratingGSS(?this) ->

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_IteratingGSS(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator implements the IteratingGSS algorithms presented in the diploma thesis 'Effiziente Entdeckung unabhängiger Subgruppen in grossen Datenbanken' at the Department of Computer Science, University of Dortmund.

iteratinggss

Parameters: hasParameterKey iterations hasDefalutValue 10 comment the number of iterations hasParameterKey force_iterations hasDefalutValue false comment make all iterations even if termination criterion is met hasParameterKey epsilon hasDefalutValue 0.04 comment approximation parameter hasParameterKey delta hasDefalutValue 0.1 comment desired confidence hasParameterKey min_utility_pruning hasDefalutValue 0.0 comment minimum utility used for pruning hasParameterKey min_utility_useful hasDefalutValue 0.0 comment minimum utility for the usefulness of a rule hasParameterKey stepsize hasDefalutValue 100 comment the number of examples drawn before the next hypothesis update hasParameterKey large hasDefalutValue 100 comment the number of examples a hypothesis must cover before normal approximation is used hasParameterKey max_complexity hasDefalutValue 1 comment the maximum complexity of hypothesis hasParameterKey min_complexity hasDefalutValue 1 comment the minimum complexity of hypothesis hasParameterKey use_binomial hasDefalutValue false comment Switch to binomial utility funtion before increasing complexity hasParameterKey utility_function hasDefalutValue wracc comment the utility function to be used hasParameterKey use_kbs hasDefalutValue true comment use kbs to reweight examples after each iteration hasParameterKey rejection_sampling hasDefalutValue true comment use rejection sampling instead of weighted examples hasParameterKey useful_criterion hasDefalutValue utility comment criterion to decide if the complexity is increased hasParameterKey example_factor hasDefalutValue 1.5 comment used by example criterion to determine usefulness of a hypothesis hasParameterKey generate_all_hypothesis hasDefalutValue false comment generate h->Y+/Y- or h->Y+ only. hasParameterKey reset_weights hasDefalutValue false comment Set weights back to 1 when complexity is increased. 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

linear_discriminant_analysis

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Linear_Discriminant_Analysis(?this)

This operator performs a linear discriminant analysis (LDA). This method tries to find the linear combination of features which best separate two or more classes of examples. The resulting combination is then used as a linear classifier. LDA is closely related to ANOVA (analysis of variance) and regression analysis, which also attempt to express one dependent variable as a linear combination of other features or measurements. In the other two methods however, the dependent variable is a numerical quantity, while for LDA it is a categorical variable (i.e. the class label). LDA is also closely related to principal component analysis (PCA) and factor analysis in that both look for linear combinations of variables which best explain the data. LDA explicitly attempts to model the difference between the classes of data. PCA on the other hand does not take into account any difference in class.

RM_Linear_Discriminant_Analysis(?this) ->

Parameters: 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

linear_regression

This operator calculates a linear regression model. It uses the Akaike criterion for model selection.

RM_Linear_Regression(?this) ->

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Linear_Regression(?this)

Parameters: hasParameterKey feature_selection hasDefalutValue M5 prime comment The feature selection method used during regression. hasParameterKey eliminate_colinear_features hasDefalutValue true comment Indicates if the algorithm should try to delete colinear features during the regression. hasParameterKey use_bias hasDefalutValue true comment Indicates if an intercept value should be calculated. hasParameterKey min_standardized_coefficient hasDefalutValue 1.5 comment The minimum standardized coefficient for the removal of colinear feature elimination. hasParameterKey ridge hasDefalutValue 1.0E-8 comment The ridge parameter used during ridge regression. 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey degree hasDefalutValue 2 comment Specifies the degree of the local fitted polynomial. Please keep in mind, that a higher degree than 2 will increase calculation time extremely and probably suffer from overfitting. hasParameterKey ridge_factor hasDefalutValue 1.0E-9 comment Specifies the ridge factor. This factor is used to penalize high coefficients. In order to aviod overfitting this might be increased. hasParameterKey use_robust_estimation hasDefalutValue false comment If checked, a reweighting of the examples is performed in order to downweight outliers hasParameterKey use_weights hasDefalutValue true comment Indicates if example weights should be used if present in the given example set. hasParameterKey iterations hasDefalutValue 20 comment The number of iterations performed for weight calculation. hasParameterKey numerical_measure hasDefalutValue EuclideanDistance comment Select measure hasParameterKey neighborhood_type hasDefalutValue Fixed Number comment Determines which type of neighborhood should be used. Either with fixed number of neighbors, or all neighbors within a distance or mixed. hasParameterKey k hasDefalutValue 5 comment Specifies the number of neighbors in the neighborhood. Regardless of the local density, always that much samples are returned. hasParameterKey fixed_distance hasDefalutValue 5.0 comment Specifies the size of the neighborhood. All points within this distance are added. hasParameterKey relative_size hasDefalutValue comment Specifies the size of the neighborhood relative to the total number of examples. A value of 0.04 would include 4% of the data points into the neighborhood. hasParameterKey distance hasDefalutValue 10.0 comment Specifies the size of the neighborhood. All points within this distance are added. hasParameterKey at_least hasDefalutValue 20 comment If the neighborhood count is less than this number, the distance is increased until this number is met. hasParameterKey smoothing_kernel hasDefalutValue Triweight comment Determines which kernel type is used to calculate the weights of distant examples.

local_polynomial_regression

This operator provides functionality to perform a local regression. That means, that if the label value for a point in the data space is requested, the local neighborhood of this point is searched. For this search the distance measure specified in the distance measure parameter is used. After the neighborhood has been determined, its datapoints are used for fitting a polynomial of the specified degree using the weighted least squares optimization. The value of this polynom at the requested point in data space is then returned as result. During the fitting of the polynom, the neighborhoods data points are weighted by their distance to the requested point. Here again the distance function specified in the parameters is used. The weight is calculated from the distance using the kernel smoother, specified in the parameters. The resulting weight is then included into the least squares optimization. If the training example set contains a weight attribute, the distance based weight is multiplied by the example's weight. If the parameter use_robust_estimation is checked, a Generate Weight (LPR) is performed with the same parameters as the following Local Polynomial Regression. For different settings the operator Generate Weight (LPR) might be used as a preprocessing step instead of checking the parameter. The effect is, that outlier will be downweighted so that the least squares fitting will not be affected by them anymore. Since it is a local method, the computational need for training is minimal: In fact, each example is only stored in a way which provides a fast neighborhood search during application time. Since all calculations are performed during application time, it is slower than for example SVM, LinearRegression or NaiveBayes. In fact it really much depends on the number of training examples and the number of attributes. If a higher degree than 1 is used, the calculations take much longer, because implicitly the polynomial expansion must be calculated.

RM_Local_Polynomial_Regression(?this) ->

[DataTable and (targetColumn exactly 1 (ScalarColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Local_Polynomial_Regression(?this)

1 1

RM_LogisticRegression(?this) ->

logisticregression

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_LogisticRegression(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey add_intercept hasDefalutValue true comment Determines whether to include an intercept. hasParameterKey start_population_type hasDefalutValue random comment The type of start population initialization. hasParameterKey max_generations hasDefalutValue 10000 comment Stop after this many evaluations hasParameterKey generations_without_improval hasDefalutValue 300 comment Stop after this number of generations without improvement (-1: optimize until max_iterations). hasParameterKey population_size hasDefalutValue 3 comment The population size (-1: number of examples) hasParameterKey tournament_fraction hasDefalutValue 0.75 comment The fraction of the population used for tournament selection. hasParameterKey keep_best hasDefalutValue true comment Indicates if the best individual should survive (elititst selection). hasParameterKey mutation_type hasDefalutValue gaussian_mutation comment The type of the mutation operator. hasParameterKey selection_type hasDefalutValue tournament comment The type of the selection operator. hasParameterKey crossover_prob hasDefalutValue 1.0 comment The probability for crossovers. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed hasParameterKey show_convergence_plot hasDefalutValue false comment Indicates if a dialog with a convergence plot should be drawn. hasParameterKey return_model_performance hasDefalutValue true comment Determines whether to return the performance. This operator determines a logistic regression model. 1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

logistic_regression

This is the Java implementation of myKLR by Stefan Rüping. myKLR is a tool for large scale kernel logistic regression based on the algorithm of Keerthi/etal/2003 and the code of mySVM.

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Logistic_Regression(?this)

RM_Logistic_Regression(?this) ->

logistic_regression_evolutionary

This operator determines a logistic regression model. Parameters: hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b. hasParameterKey C hasDefalutValue 1.0 comment The complexity constant. hasParameterKey start_population_type hasDefalutValue random comment The type of start population initialization. hasParameterKey max_generations hasDefalutValue 10000 comment Stop after this many evaluations hasParameterKey generations_without_improval hasDefalutValue 30 comment Stop after this number of generations without improvement (-1: optimize until max_iterations). hasParameterKey population_size hasDefalutValue 1 comment The population size (-1: number of examples) hasParameterKey tournament_fraction hasDefalutValue 0.75 comment The fraction of the population used for tournament selection. hasParameterKey keep_best hasDefalutValue true comment Indicates if the best individual should survive (elititst selection). hasParameterKey mutation_type hasDefalutValue gaussian_mutation comment The type of the mutation operator. hasParameterKey selection_type hasDefalutValue tournament comment The type of the selection operator. hasParameterKey crossover_prob hasDefalutValue 1.0 comment The probability for crossovers. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed hasParameterKey show_convergence_plot hasDefalutValue false comment Indicates if a dialog with a convergence plot should be drawn.

RM_Logistic_Regression_Evolutionary(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Logistic_Regression_Evolutionary(?this)

1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_MinMaxBinDiscretization_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

RM_MinMaxBinDiscretization_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

minmaxbindiscretization

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_MinMaxBinDiscretization_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_), meanValue(?Column, ?_), minimumValue(?Column, ?_), maximumValue(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_MinMaxBinDiscretization_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

minmaxbindiscretization

This operator discretizes all numeric attributes in the dataset into nominal attributes. This discretization is performed by simple binning, i.e. the specified number of equally sized bins is created and the numerical values are simply sorted into those bins. Skips all special attributes including the label. In contrast to the usual simple binning performed by the BinDiscretization, this operator bins the values into a predefined range (and not into the range defined by the minimum and maximum values taken from the data). Parameters: hasParameterKey return_preprocessing_model hasDefalutValue false comment Indicates if the preprocessing model should also be returned hasParameterKey create_view hasDefalutValue false comment Create View to apply preprocessing instead of changing the data hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey number_of_bins hasDefalutValue 2 comment Defines the number of bins which should be used for each attribute. hasParameterKey min_value hasDefalutValue comment The minimum value for the binning range. hasParameterKey max_value hasDefalutValue comment The maximum value for the binning range. hasParameterKey range_name_type hasDefalutValue long comment Indicates if long range names including the limits should be used. hasParameterKey automatic_number_of_digits hasDefalutValue true comment Indicates if the number of digits should be automatically determined for the range names. hasParameterKey number_of_digits hasDefalutValue -1 comment The minimum number of digits used for the interval names (-1: determine minimal number automatically). 1 1

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_MultiCriterionDecisionStump(?this)

multicriteriondecisionstump

RM_MultiCriterionDecisionStump(?this) ->

Parameters: hasParameterKey utility_function hasDefalutValue entropy comment The function to be optimized by the rule.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

A DecisionStump clone that allows to specify different utility functions. It is quick for nominal attributes, but does not yet apply pruning for continuos attributes. Currently it can only handle boolean class labels. 1 1

naive_bayes

Naive Bayes learner.

RM_Naive_Bayes(?this) ->

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_Naive_Bayes(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey laplace_correction hasDefalutValue true comment Use Laplace correction to prevent high influence of zero probabilities. 1 1

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_Naive_Bayes_Kernel(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey laplace_correction hasDefalutValue true comment Use Laplace correction to prevent high influence of zero probabilities. hasParameterKey estimation_mode hasDefalutValue greedy comment The kernel density estimation mode. hasParameterKey bandwidth_selection hasDefalutValue heuristic comment The method to set the kernel bandwidth. hasParameterKey bandwidth hasDefalutValue 0.1 comment Kernel bandwidth. hasParameterKey minimum_bandwidth hasDefalutValue 0.1 comment Minimum kernel bandwidth. hasParameterKey number_of_kernels hasDefalutValue 10 comment Number of kernels. hasParameterKey use_application_grid hasDefalutValue false comment Use a kernel density function grid in model application. (Speeds up model application at the expense of the density function precision.) hasParameterKey application_grid_size hasDefalutValue 200 comment Size of the application grid. Kernel Naive Bayes learner.

naive_bayes_kernel

RM_Naive_Bayes_Kernel(?this) ->

1 1

neuralnet

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_NeuralNet(?this) ->

Parameters: hasParameterKey input_layer_type hasDefalutValue linear comment The default layer type for the input layers. hasParameterKey output_layer_type hasDefalutValue sigmoid comment The default layer type for the output layers. hasParameterKey default_number_of_hidden_layers hasDefalutValue 1 comment The number of hidden layers. Only used if no layers are defined by the list hidden_layer_types. hasParameterKey default_hidden_layer_size hasDefalutValue -1 comment The default size of hidden layers. Only used if no layers are defined by the list hidden_layer_types. -1 means size (number of attributes + number of classes) / 2 hasParameterKey default_hidden_layer_type hasDefalutValue sigmoid comment The default layer type for the hidden layers. Only used if the parameter list hidden_layer_types is not defined. hasParameterKey hidden_layer_types hasDefalutValue comment Describes the type and size of all hidden layers hasParameterKey training_cycles hasDefalutValue 200 comment The number of training cycles used for the neural network training. hasParameterKey learning_rate hasDefalutValue 0.3 comment The learning rate determines by how much we change the weights at each step. hasParameterKey momentum hasDefalutValue 0.2 comment The momentum simply adds a fraction of the previous weight update to the current one (prevent local maxima and smoothes optimization directions). hasParameterKey error_epsilon hasDefalutValue 0.05 comment The optimization is stopped if the training error gets below this epsilon value. This operator learns a model by means of a feed-forward neural network. The learning is done via backpropagation. The user can define the structure of the neural network with the parameter list "hidden_layer_types". Each list entry describes a new hidden layer. The key of each entry must correspond to the layer type which must be one out of <ul> <li>linear</li> <li>sigmoid (default)</li> <li>tanh</li> <li>sine</li> <li>logarithmic</li> <li>gaussian</li> </ul> The key of each entry must be a number defining the size of the hidden layer. A size value of -1 or 0 indicates that the layer size should be calculated from the number of attributes of the input example set. In this case, the layer size will be set to (number of attributes + number of classes) / 2 + 1. If the user does not specify any hidden layers, a default hidden layer with sigmoid type and size (number of attributes + number of classes) / 2 + 1 will be created and added to the net. The type of the input nodes is sigmoid. The type of the output node is sigmoid is the learning data describes a classification task and linear for numerical regression tasks.

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_NeuralNet(?this)

1 1 This operator learns a model by means of a feed-forward neural network. The learning is done via backpropagation. The user can define the structure of the neural network in two different ways according to the setting of the parameter define_different_hidden_layers. If different hidden layers are defined, the parameter hidden_layer_sizes must be set to a comma separated list of the sizes of all hidden layers, e.g. 3,7,5. If no different hidden layers are defined, the parameters for the default hidden layers are used. A size value of -1 or 0 indicates that the layer size should be calculated from the number of attributes of the input example set. In this case, the layer size will be set to (number of attributes + number of classes) / 2 + 1. All layers have a sigmoid activation function. If the user does not specify any hidden layers, a default hidden layer with size (number of attributes + number of classes) / 2 + 1 will be created and added to the net.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

neuralnetsimple

RM_NeuralNetSimple(?this) ->

Parameters: hasParameterKey default_number_of_hidden_layers hasDefalutValue 1 comment The number of hidden layers. Only used if no layers are defined by the list hidden_layer_types. hasParameterKey default_hidden_layer_size hasDefalutValue -1 comment The default size of hidden layers. Only used if no layers are defined by the list hidden_layer_types. -1 means size (number of attributes + number of classes) / 2 hasParameterKey training_cycles hasDefalutValue 500 comment The number of training cycles used for the neural network training. hasParameterKey learning_rate hasDefalutValue 0.3 comment The learning rate determines by how much we change the weights at each step. hasParameterKey momentum hasDefalutValue 0.2 comment The momentum simply adds a fraction of the previous weight update to the current one (prevent local maxima and smoothes optimization directions). hasParameterKey error_epsilon hasDefalutValue 0.01 comment The optimization is stopped if the training error gets below this epsilon value. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_NeuralNetSimple(?this)

1 1 This operator learns a model by means of a feed-forward neural network trained by a backpropagation algorithm (multi-layer perceptron). The user can define the structure of the neural network with the parameter list "hidden_layers". Each list entry describes a new hidden layer. The key of each entry must correspond to the layer name. The value of each entry must be a number defining the size of the hidden layer. A size value of -1 indicates that the layer size should be calculated from the number of attributes of the input example set. In this case, the layer size will be set to (number of attributes + number of classes) / 2 + 1. If the user does not specify any hidden layers, a default hidden layer with sigmoid type and size (number of attributes + number of classes) / 2 + 1 will be created and added to the net. If only a single layer without nodes is specified, the input nodes are directly connected to the output nodes and no hidden layer will be used. The used activation function is the usual sigmoid function. Therefore, the values ranges of the attributes should be scaled to -1 and +1. This is also done by this operator if not specified otherwise by the corresponding parameter setting. The type of the output node is sigmoid if the learning data describes a classification task and linear for numerical regression tasks. Parameters: hasParameterKey hidden_layers hasDefalutValue comment Describes the name and the size of all hidden layers. hasParameterKey training_cycles hasDefalutValue 500 comment The number of training cycles used for the neural network training. hasParameterKey learning_rate hasDefalutValue 0.3 comment The learning rate determines by how much we change the weights at each step. hasParameterKey momentum hasDefalutValue 0.2 comment The momentum simply adds a fraction of the previous weight update to the current one (prevent local maxima and smoothes optimization directions). hasParameterKey decay hasDefalutValue false comment Indicates if the learning rate should be decreased during learningh hasParameterKey shuffle hasDefalutValue true comment Indicates if the input data should be shuffled before learning (increases memory usage but is recommended if data is sorted before) hasParameterKey normalize hasDefalutValue true comment Indicates if the input data should be normalized between -1 and +1 before learning (increases runtime but is in most cases necessary) hasParameterKey error_epsilon hasDefalutValue 1.0E-5 comment The optimization is stopped if the training error gets below this epsilon value. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Neural_Net(?this) ->

neural_net

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Neural_Net(?this)

1 1 1

CategorialColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Nominal_to_Numerical_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

nominal_to_numerical

RM_Nominal_to_Numerical_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), CategorialColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iScalar), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

This operator maps all non numeric attributes to real valued attributes. Nothing is done for numeric attributes, binary attributes are mapped to 0 and 1. For nominal attributes one of the following calculations will be done: <ul> <li>Dichotomization, i.e. one new attribute for each value of the nominal attribute. The new attribute which corresponds to the actual nominal value gets value 1 and all other attributes gets value 0.</li> <li>Alternatively the values of nominal attributes can be seen as equally ranked, therefore the nominal attribute will simply be turned into a real valued attribute, the old values results in equidistant real values.</li> </ul> At this moment the same applies for ordinal attributes, in a future release more appropriate values based on the ranking between the ordinal values may be included. 1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Nominal_to_Numerical_single(?this), uses(?this, ?Din), containsColumn(?Din,?Column), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), amountOfMissingValues(?Column, ?OldMissingCount), amountOfDifferentValues(?Column,?AV), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), newFor(?NewColumn, ?Column), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), representsAttribute(?NewColumn, ?Attribute), columnHasType(?NewColumn, iScalar), minimumValue(?NewColumn,1), maximumValue(?NewColumn, ?AV), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

CategorialColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Nominal_to_Numerical_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

nominal_to_numerical

normalize

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Normalize_all_Z_transformation(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), not({ meanValue(?Column,0), standardDeviation(?Column,1) }), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all"), simpleParameter_method(?this, "Z-transformation") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), newFor(?NewColumn, ?Column), representsAttribute(?NewColumn, ?Attribute), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iScalar), meanValue(?NewColumn, 0), standardDeviation(?NewColumn, 1), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

ScalarColumn(?Column), not({ meanValue(?Column,0), standardDeviation(?Column,1) }), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Normalize_all_Z_transformation(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all"), simpleParameter_method(?this, "Z-transformation")

1 1 1

ScalarColumn(?Column), not({ minimumValue(?Column,0), maximumValue(?Column,1) }), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Normalize_all_range_transformation(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all"), simpleParameter_method(?this, "range transformation")

RM_Normalize_all_range_transformation(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), simpleParameter_attribute_filter_type(?this, "all"), simpleParameter_method(?this, "range transformation") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), newFor(?NewColumn, ?Column), representsAttribute(?NewColumn, ?Attribute), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iScalar), minimumValue(?NewColumn, 0), maximumValue(?NewColumn, 1), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator performs a normalization. This can be done between a user defined minimum and maximum value or by a z-transformation, i.e. on mean 0 and variance 1. or by a proportional transformation as proportion of the total sum of the respective attribute.

normalize

1 1 1

RM_Normalize_single_Z_transformation(?this), uses(?this, ?Din), containsColumn(?Din,?Column), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single"), simpleParameter_method(?this, "Z-transformation") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), newFor(?NewColumn, ?Column), representsAttribute(?NewColumn, ?Attribute), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iScalar), meanValue(?NewColumn, 0), standardDeviation(?NewColumn, 1), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

ScalarColumn(?Column), not({ meanValue(?Column,0), standardDeviation(?Column,1) }), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Normalize_single_Z_transformation(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single"), simpleParameter_method(?this, "Z-transformation")

normalize

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey return_preprocessing_model hasDefalutValue false comment Indicates if the preprocessing model should also be returned hasParameterKey create_view hasDefalutValue false comment Create View to apply preprocessing instead of changing the data hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey method hasDefalutValue Z-transformation comment Select the normalization method. hasParameterKey min hasDefalutValue 0.0 comment The minimum value after normalization hasParameterKey max hasDefalutValue 1.0 comment The maximum value after normalization 1 1 1 Parameters: hasParameterKey return_preprocessing_model hasDefalutValue false comment Indicates if the preprocessing model should also be returned hasParameterKey create_view hasDefalutValue false comment Create View to apply preprocessing instead of changing the data hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey method hasDefalutValue Z-transformation comment Select the normalization method. hasParameterKey min hasDefalutValue 0.0 comment The minimum value after normalization hasParameterKey max hasDefalutValue 1.0 comment The maximum value after normalization

ScalarColumn(?Column), not({ minimumValue(?Column,0), maximumValue(?Column,1) }), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Normalize_single_range_transformation(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single"), simpleParameter_method(?this, "range transformation")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Normalize_single_range_transformation(?this), uses(?this, ?Din), containsColumn(?Din,?Column), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single"), simpleParameter_method(?this, "range transformation") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), newFor(?NewColumn, ?Column), representsAttribute(?NewColumn, ?Attribute), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iScalar), minimumValue(?NewColumn, 0), maximumValue(?NewColumn, 1), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

normalize

1 1 1 Parameters: hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey min hasDefalutValue 0.0 comment The minimal value which is mapped to false (included). hasParameterKey max hasDefalutValue 0.0 comment The maximal value which is mapped to false (included).

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Numerical_to_Binominal_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iBinary), minimumValue(?NewColumn, 0), maximumValue(?NewColumn, 1), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

Converts all numerical attributes to binary ones. If the value of an attribute is between the specified minimal and maximal value, it becomes false, otherwise true. If the value is missing, the new value will be missing. The default boundaries are both set to 0, thus only 0.0 is mapped to false and all other values are mapped to true.

numerical_to_binominal

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Numerical_to_Binominal_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

1 1 1

numerical_to_binominal

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Numerical_to_Binominal_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

Parameters: hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue numeric comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue real comment Except this value type. hasParameterKey block_type hasDefalutValue value_series comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_series_end comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept. hasParameterKey min hasDefalutValue 0.0 comment The minimal value which is mapped to false (included). hasParameterKey max hasDefalutValue 0.0 comment The maximal value which is mapped to false (included).

RM_Numerical_to_Binominal_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iBinary), minimumValue(?NewColumn, 0), maximumValue(?NewColumn, 1), amountOfMissingValues(?NewColumn, ?OldMissingCount), produces(?this, ?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Numerical_to_Polynominal_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), ScalarColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

Converts all numerical attributes to nominal ones. Each numerical value is simply used as nominal value of the new attribute. If the value is missing, the new value will be missing. Please note that this operator might drastically increase memory usage if many different numerical values are used. Please use the available discretization operators then.

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Numerical_to_Polynominal_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

numerical_to_polynominal

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

numerical_to_polynominal

RM_Numerical_to_Polynominal_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), columnHasType(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), columnHasType(?NewColumn, iCategorial), produces(?this, ?Dout)

ScalarColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Numerical_to_Polynominal_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

1 1

RM_Perceptron(?this) ->

The perceptron is a type of artificial neural network invented in 1957 by Frank Rosenblatt. It can be seen as the simplest kind of feedforward neural network: a linear classifier. Beside all biological analogies, the single layer perceptron is simply a linear classifier which is efficiently trained by a simple update rule: for all wrongly classified data points, the weight vector is either increased or decreased by the corresponding example values.

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Perceptron(?this)

perceptron

Parameters: hasParameterKey rounds hasDefalutValue 3 comment The number of datascans used to adapt the hyperplane. hasParameterKey learning_rate hasDefalutValue 0.05 comment The hyperplane will adapt with this rate to each example.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1

[DataTable and (targetColumn exactly 1 (ScalarColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Polynomial_Regression(?this)

RM_Polynomial_Regression(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey max_iterations hasDefalutValue 5000 comment The maximum number of iterations used for model fitting. hasParameterKey replication_factor hasDefalutValue 1 comment The amount of times each input variable is replicated, i.e. how many different degrees and coefficients can be applied to each variable hasParameterKey max_degree hasDefalutValue 5 comment The maximal degree used for the final polynomial. hasParameterKey min_coefficient hasDefalutValue -100.0 comment The minimum number used for the coefficients and the offset. hasParameterKey max_coefficient hasDefalutValue 100.0 comment The maximum number used for the coefficients and the offset. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

polynomial_regression

This regression learning operator fits a polynomial of all attributes to the given data set. If the data set contains a label Y and three attributes X1, X2, and X3 a function of the form <code>Y = w0 + w1 * X1 ^ d1 + w2 * X2 ^ d2 + w3 * X3 ^ d3</code> will be fitted to the training data. 1 1 This operator performs a quadratic discriminant analysis (QDA). QDA is closely related to linear discriminant analysis (LDA), where it is assumed that the measurements are normally distributed. Unlike LDA however, in QDA there is no assumption that the covariance of each of the classes is identical.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

quadratic_discriminant_analysis

RM_Quadratic_Discriminant_Analysis(?this) ->

Parameters:

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Quadratic_Discriminant_Analysis(?this)

1 1

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_Random_Forest(?this)

Parameters: hasParameterKey number_of_trees hasDefalutValue 10 comment The number of learned random trees. hasParameterKey criterion hasDefalutValue gain_ratio comment Specifies the used criterion for selecting attributes and numerical splits. hasParameterKey minimal_size_for_split hasDefalutValue 4 comment The minimal size of a node in order to allow a split. hasParameterKey minimal_leaf_size hasDefalutValue 2 comment The minimal size of all leaves. hasParameterKey minimal_gain hasDefalutValue 0.1 comment The minimal gain which must be achieved in order to produce a split. hasParameterKey maximal_depth hasDefalutValue 20 comment The maximum tree depth (-1: no bound) hasParameterKey confidence hasDefalutValue 0.25 comment The confidence level used for the pessimistic error calculation of pruning. hasParameterKey number_of_prepruning_alternatives hasDefalutValue 3 comment The number of alternative nodes tried when prepruning would prevent a split. hasParameterKey no_pre_pruning hasDefalutValue false comment Disables the pre pruning and delivers a tree without any prepruning. hasParameterKey no_pruning hasDefalutValue false comment Disables the pruning and delivers an unpruned tree. hasParameterKey guess_subset_ratio hasDefalutValue true comment Indicates that log(m) + 1 features are used, otherwise a ratio has to be specified. hasParameterKey subset_ratio hasDefalutValue 0.2 comment Ratio of randomly chosen attributes to test hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

random_forest

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Random_Forest(?this) ->

This operators learns a random forest. The resulting forest model contains serveral single random tree models. 1 1

random_tree

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_Random_Tree(?this)

RM_Random_Tree(?this) ->

Parameters: hasParameterKey criterion hasDefalutValue gain_ratio comment Specifies the used criterion for selecting attributes and numerical splits. hasParameterKey minimal_size_for_split hasDefalutValue 4 comment The minimal size of a node in order to allow a split. hasParameterKey minimal_leaf_size hasDefalutValue 2 comment The minimal size of all leaves. hasParameterKey minimal_gain hasDefalutValue 0.1 comment The minimal gain which must be achieved in order to produce a split. hasParameterKey maximal_depth hasDefalutValue 20 comment The maximum tree depth (-1: no bound) hasParameterKey confidence hasDefalutValue 0.25 comment The confidence level used for the pessimistic error calculation of pruning. hasParameterKey number_of_prepruning_alternatives hasDefalutValue 3 comment The number of alternative nodes tried when prepruning would prevent a split. hasParameterKey no_pre_pruning hasDefalutValue false comment Disables the pre pruning and delivers a tree without any prepruning. hasParameterKey no_pruning hasDefalutValue false comment Disables the pruning and delivers an unpruned tree. hasParameterKey guess_subset_ratio hasDefalutValue true comment Indicates that log(m) + 1 features are used, otherwise a ratio has to be specified. hasParameterKey subset_ratio hasDefalutValue 0.2 comment Ratio of randomly chosen attributes to test hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed This operator learns decision trees from both nominal and numerical data. Decision trees are powerful classification methods which often can also easily be understood. The random tree learner works similar to Quinlan's C4.5 or CART but it selects a random subset of attributes before it is applied. The size of the subset is defined by the parameter subset_ratio. 1 1 Parameters: hasParameterKey alpha hasDefalutValue 0.5 comment This is the strength of regularization. 1: Only global covariance is used, 0: Only per class covariance is used.

regularized_discriminant_analysis

This is a regularized discriminant analysis (RDA) which is a generalization of the LDA or QDA. Both algorithms are special cases of this algorithm, using parameter alpha = 1 respectively alpha = 0.

RM_Regularized_Discriminant_Analysis(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Regularized_Discriminant_Analysis(?this)

1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

relevance_vector_machine

Parameters: hasParameterKey rvm_type hasDefalutValue Regression-RVM comment Regression RVM hasParameterKey kernel_type hasDefalutValue rbf comment The type of the kernel functions. hasParameterKey kernel_lengthscale hasDefalutValue 3.0 comment The lengthscale used in all kernels. hasParameterKey kernel_degree hasDefalutValue 2.0 comment The degree used in the poly kernel. hasParameterKey kernel_bias hasDefalutValue 1.0 comment The bias used in the poly kernel. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The SVM kernel parameter sigma1 (Epanechnikov, Gaussian Combination, Multiquadric). hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The SVM kernel parameter sigma2 (Gaussian Combination). hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The SVM kernel parameter sigma3 (Gaussian Combination). hasParameterKey kernel_shift hasDefalutValue 1.0 comment The SVM kernel parameter shift (Multiquadric). hasParameterKey kernel_a hasDefalutValue 1.0 comment The SVM kernel parameter a (neural). hasParameterKey kernel_b hasDefalutValue 0.0 comment The SVM kernel parameter b (neural). hasParameterKey max_iteration hasDefalutValue 100 comment The maximum number of iterations used. hasParameterKey min_delta_log_alpha hasDefalutValue 0.0010 comment Abort iteration if largest log alpha change is smaller than this hasParameterKey alpha_max hasDefalutValue 1.0E12 comment Prune basis function if its alpha is bigger than this hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed Relevance Vector Machine (RVM) Learner. The RVM is a probabilistic method both for classification and regression. The implementation of the relevance vector machine is based on the original algorithm described by Tipping/2001. The fast version of the marginal likelihood maximization (Tipping/Faul/2003) is also available if the parameter "rvm_type" is set to "Constructive-Regression-RVM".

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Relevance_Vector_Machine(?this)

RM_Relevance_Vector_Machine(?this) ->

1 1 1

replace_missing_values

RM_Replace_Missing_Values_all(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), representsAttribute(?Column, ?Attribute), NonMissingValueFreeColumn(?Column), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "all") -> copy(?Dout,?Din,{containsColumn(?Din, ?_)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), amountOfMissingValues(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), amountOfMissingValues(?NewColumn, 0), produces(?this, ?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Replaces missing values in examples. If a value is missing, it is replaced by one of the functions "minimum", "maximum", "average", and "none", which is applied to the non missing attribute values of the example set. "none" means, that the value is not replaced. The function can be selected using the parameter list <code>columns</code>. If an attribute's name appears in this list as a key, the value is used as the function name. If the attribute's name is not in the list, the function specified by the <code>default</code> parameter is used. For nominal attributes the mode is used for the average, i.e. the nominal value which occurs most often in the data. For nominal attributes and replacement type zero the first nominal value defined for this attribute is used. The replenishment "value" indicates that the user defined parameter should be used for the replacement.

NonMissingValueFreeColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Replace_Missing_Values_all(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "all")

1 1 1

NonMissingValueFreeColumn(?Column), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), containsColumn(?Din,?Column) -> RM_Replace_Missing_Values_single(?this), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute(?this, ?AttName), simpleParameter_attribute_filter_type(?this, "single")

RM_Replace_Missing_Values_single(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute, ?AttName), representsAttribute(?Column, ?Attribute), amountOfMissingValues(?Column, ?OldMissingCount), representsAttribute(?Column, ?Attribute), simpleParameter_attribute_filter_type(?this, "single") -> copy(?Dout,?Din,{containsColumn(?Din, ?Column)}), copy(?NewColumn, ?Column, { complexObjectPart(?_, ?Column), amountOfMissingValues(?Column, ?_)}), DataColumn(?NewColumn), containsColumn(?Dout, ?NewColumn), amountOfMissingValues(?NewColumn, 0), produces(?this, ?Dout)

replace_missing_values

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Rule_Induction(?this) ->

Parameters: hasParameterKey criterion hasDefalutValue information_gain comment Specifies the used criterion for selecting attributes and numerical splits. hasParameterKey sample_ratio hasDefalutValue 0.9 comment The sample ratio of training data used for growing and pruning. hasParameterKey pureness hasDefalutValue 0.9 comment The desired pureness, i.e. the necessary amount of the major class in a covered subset in order become pure. hasParameterKey minimal_prune_benefit hasDefalutValue 0.25 comment The minimum amount of benefit which must be exceeded over unpruned benefit in order to be pruned. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn))](?D) -> RM_Rule_Induction(?this)

This operator works similar to the propositional rule learner named Repeated Incremental Pruning to Produce Error Reduction (RIPPER, Cohen 1995). Starting with the less prevalent classes, the algorithm iteratively grows and prunes rules until there are no positive examples left or the error rate is greater than 50%. In the growing phase, for each rule greedily conditions are added to the rule until the rule is perfect (i.e. 100% accurate). The procedure tries every possible value of each attribute and selects the condition with highest information gain. In the prune phase, for each rule any final sequences of the antecedents is pruned with the pruning metric p/(p+n).

rule_induction

1 1

amountOfRows(?Din, ?NumRows) -> RM_Sample_Bootstrapping(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

sample_bootstrapping

Parameters: hasParameterKey sample hasDefalutValue relative comment Determines how the amount of data is specified. hasParameterKey sample_size hasDefalutValue 100 comment The number of examples which should be sampled hasParameterKey sample_ratio hasDefalutValue 1.0 comment This ratio determines the size of the new example set. hasParameterKey use_weights hasDefalutValue true comment If checked, example weights will be considered during the bootstrapping if such weights are present. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed This operator constructs a bootstrapped sample from the given example set. That means that a sampling with replacement will be performed. The usual sample size is the number of original examples. This operator also offers the possibility to create the inverse example set, i.e. an example set containing all examples which are not part of the bootstrapped example set. This inverse example set might be used for a bootstrapped validation (together with an IteratingPerformanceAverage operator.

RM_Sample_Bootstrapping(?this), uses(?this, ?Din) -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), produces(?this, ?Dout)

1 1

RM_Sample_Kennard_Stone(?this), uses(?this, ?Din) -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), produces(?this, ?Dout)

amountOfRows(?Din, ?NumRows) -> RM_Sample_Kennard_Stone(?this)

This operator performs a Kennard-Stone Sampling. This sampling Algorithm works as follows: First find the two points most separated in the training set. For each candidate point, find the smallest distance to any object already selected. Select that point for the training set which has the largest of these smallest distances As described above, this algorithm always gives the same result, due to the two starting points which are always the same. This implementation reduces number of iterations by holding a list with candidates of the largest smallest distances. The parameters controll the number of examples in the sample Parameters: hasParameterKey sample hasDefalutValue absolute comment Determines how the amount of data is specified. hasParameterKey sample_size hasDefalutValue 100 comment The number of examples which should be sampled hasParameterKey sample_ratio hasDefalutValue 0.1 comment The fraction of examples which should be sampled

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

sample_kennard_stone

1 1 1

sample_model_based

RM_Sample_Model_Based(?this), uses(?this, ?Din) -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), produces(?this, ?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

amountOfRows(?Din, ?NumRows) -> RM_Sample_Model_Based(?this)

1 1

amountOfRows(?Din, ?NumRows) -> RM_Sample_Stratified(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey sample hasDefalutValue absolute comment Determines how the amount of data is specified. hasParameterKey sample_size hasDefalutValue 100 comment The number of examples which should be sampled hasParameterKey sample_ratio hasDefalutValue 0.1 comment The fraction of examples which should be sampled hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed Stratified sampling operator. This operator performs a random sampling of a given fraction. In contrast to the simple sampling operator, this operator performs a stratified sampling for data sets with nominal label attributes, i.e. the class distributions remains (almost) the same after sampling. Hence, this operator cannot be applied on data sets without a label or with a numerical label. In these cases a simple sampling without stratification is performed.

sample_stratified

RM_Sample_Stratified(?this), uses(?this, ?Din) -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), produces(?this, ?Dout)

1 1

amountOfRows(?Din, ?NumRows), integerDivide(?NewRowCount, ?NumRows, 10) -> RM_Sample_probability_01(?this), simpleParameter_sample(?this, "probability"), simpleParameter_sample_probability(?this, "01")

RM_Sample_probability_01(?this), uses(?this, ?Din), simpleParameter_sample(?this, "probability"), simpleParameter_sample_probability(?this, "01") -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), amountOfRows(?Dout, ?NewRowCount), produces(?this, ?Dout)

Simple sampling operator. This operator performs a random sampling of a given fraction. For example, if the input example set contains 5000 examples and the sample ratio is set to 0.1, the result will have approximately 500 examples.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

sample

RM_Sample_relative_01(?this), uses(?this, ?Din), simpleParameter_sample(?this, "relative"), simpleParameter_sample_ratio(?this, "01") -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), amountOfRows(?Dout, ?NewRowCount), produces(?this, ?Dout)

sample

amountOfRows(?Din, ?NumRows), integerDivide(?NewRowCount, ?NumRows, 10) -> RM_Sample_relative_01(?this), simpleParameter_sample(?this, "relative"), simpleParameter_sample_ratio(?this, "01")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Select_Attributes_single(?this), uses(?this, ?Din), containsColumn(?Din, ?Column), hasAttribute(?Din, ?Attribute), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName), simpleParameter_attribute_filter_type(?this, "single") -> simpleParameter_attribute(?this,?AttName), copy(?Dout,?Din,{containsColumn(?Din, ?Column), hasAttribute(?Din,?Attribute)}), produces(?this,?Dout)

containsColumn(?Din, ?Column), hasAttribute(?Din, ?Attribute), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName) -> RM_Select_Attributes_single(?this), simpleParameter_attribute(?this,?AttName), simpleParameter_invert_selection(?this, 1), simpleParameter_include_special_attributes(?this, 1), simpleParameter_attribute_filter_type(?this, "single")

This operator selects which attributes of an ExampleSet should be kept and which are removed. Therefore, different filter types may be selected in the parameter attribute filter type and only attributes fulfilling this condition type are selected. The rest will be removed from the ExampleSet. There's a global switch to invert the outcome, so that all attributes which would have been originally discarded will be kept and vice versa. To invert the decision, use the invert selection parameter. These types are available <ul> <li> all: Will simply select each attribute </li> <li> single: This will allow you to select a single attribute name. It might be selected from the drop down box of parameter attribute if the meta data is known </li> <li> subset: Let's you choose a number of attributes from a list. This will not work if no meta data is present. Each known attribute is shown in the list and might be selected. </li> <li> regular_expression: This let's you specify a regular expression. Each attribute whose name matches this expression will be selected. Regular expressions are a very powerful tool but need a detailed explanation to beginners. Please refer to one of the several tutorials available on the internet for a more detailed description. </li> <li> value_type: Select only attributes of a certain type. Please mention that the types are hierarchical: For example are binominal attributes nomina as well as polynominal. </li> <li> block_type: Similar to value_type this let's you select the attributes depending on their block type. </li> <li> no_missing_values: Will select all attributes which don't contain a missing value in any example. </li> <li> numeric_value_filter: This will select the attributes by testing if all their values of all examples match this condition or if they aren't not numerical. The numeric condition might be specified by typing a numerical condition. For example the parameter string "> 6" will keep all nominal attributes and all numeric attributes having a value of greater 6 in every example. A combination of conditions is possible: "> 6 && < 11" or "<= 5 || < 0". But && and || must not be mixed. </li> </ul>

select_attributes

Parameters: hasParameterKey attribute_filter_type hasDefalutValue all comment The condition specifies which attributes are selected or affected by this operator. hasParameterKey attribute hasDefalutValue comment The attribute which should be chosen. hasParameterKey attributes hasDefalutValue comment The attribute which should be chosen. hasParameterKey regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be kept. hasParameterKey use_except_expression hasDefalutValue false comment If enabled, an exception to the specified regular expression might be specified. Attributes of matching this will be filtered out, although matching the first expression. hasParameterKey except_regular_expression hasDefalutValue comment A regular expression for the names of the attributes which should be filtered out although matching the above regular expression. hasParameterKey value_type hasDefalutValue attribute_value comment The value type of the attributes. hasParameterKey use_value_type_exception hasDefalutValue false comment If enabled, an exception to the specified value type might be specified. Attributes of this type will be filtered out, although matching the first specified type. hasParameterKey except_value_type hasDefalutValue time comment Except this value type. hasParameterKey block_type hasDefalutValue attribute_block comment The block type of the attributes. hasParameterKey use_block_type_exception hasDefalutValue false comment If enabled, an exception to the specified block type might be specified. hasParameterKey except_block_type hasDefalutValue value_matrix_row_start comment Except this block type. hasParameterKey numeric_condition hasDefalutValue comment Parameter string for the condition, e.g. '>= 5' hasParameterKey invert_selection hasDefalutValue false comment Indicates if only attributes should be accepted which would normally filtered. hasParameterKey include_special_attributes hasDefalutValue false comment Indicate if this operator should also be applied on the special attributes. Otherwise they are always kept.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 Parameters: hasParameterKey name hasDefalutValue comment The name of the attribute whose role should be changed. hasParameterKey target_role hasDefalutValue regular comment The target role of the attribute (only changed if parameter change_attribute_type is true).

set_role

RM_Set_Role_id(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute,?AttName), representsAttribute(?Column,?Attribute), simpleParameter_target_role(?this, "id") -> copy(?Dout,?Din,{hasAttribute(?Din, ?Attribute)}), identifier(?Dout, ?Attribute), produces(?this,?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator can be used to change the role of an attribute of the input ExampleSet. If you want to change the attribute name you should use the <a href="rm://opdoc/rename">Rename</a> operator. The target role indicates if the attribute is a regular attribute (used by learning operators) or a special attribute (e.g. a label or id attribute). The following target attribute types are possible: <ul> <li> regular: only regular attributes are used as input variables for learning tasks </li> <li> id: the id attribute for the example set </li> <li> label: target attribute for learning </li> <li> prediction: predicted attribute, i.e. the predictions of a learning scheme </li> <li> cluster: indicates the membership to a cluster </li> <li> weight: indicates the weight of the example </li> <li> batch: indicates the membership to an example batch </li> </ul> Users can also define own attribute types by simply using the desired name. Please be aware that roles have to be unique! Assigning a non regular role the second time will cause the first attribute to be dropped from the example set. If you want to keep this attribute, you have to change it's role first.

containsColumn(?Din, ?Column), hasAttribute(?Din, ?Attribute), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName) -> RM_Set_Role_id(?this), simpleParameter_attribute(?this, ?AttName), simpleParameter_include_special_attributes(?this, 1), simpleParameter_target_role(?this, "id")

1 1

containsColumn(?Din, ?Column), hasAttribute(?Din, ?Attribute), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName) -> RM_Set_Role_label(?this), simpleParameter_attribute(?this, ?AttName), simpleParameter_include_special_attributes(?this, 1), simpleParameter_target_role(?this, "label")

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Set_Role_label(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute,?AttName), representsAttribute(?Column,?Attribute), simpleParameter_target_role(?this, "label") -> copy(?Dout,?Din,{hasAttribute(?Din, ?Attribute)}), targetAttribute(?Dout, ?Attribute), produces(?this,?Dout)

Parameters: hasParameterKey name hasDefalutValue comment The name of the attribute whose role should be changed. hasParameterKey target_role hasDefalutValue regular comment The target role of the attribute (only changed if parameter change_attribute_type is true). This operator can be used to change the role of an attribute of the input ExampleSet. If you want to change the attribute name you should use the <a href="rm://opdoc/rename">Rename</a> operator. The target role indicates if the attribute is a regular attribute (used by learning operators) or a special attribute (e.g. a label or id attribute). The following target attribute types are possible: <ul> <li> regular: only regular attributes are used as input variables for learning tasks </li> <li> id: the id attribute for the example set </li> <li> label: target attribute for learning </li> <li> prediction: predicted attribute, i.e. the predictions of a learning scheme </li> <li> cluster: indicates the membership to a cluster </li> <li> weight: indicates the weight of the example </li> <li> batch: indicates the membership to an example batch </li> </ul> Users can also define own attribute types by simply using the desired name. Please be aware that roles have to be unique! Assigning a non regular role the second time will cause the first attribute to be dropped from the example set. If you want to keep this attribute, you have to change it's role first.

set_role

1 1 This operator can be used to change the role of an attribute of the input ExampleSet. If you want to change the attribute name you should use the <a href="rm://opdoc/rename">Rename</a> operator. The target role indicates if the attribute is a regular attribute (used by learning operators) or a special attribute (e.g. a label or id attribute). The following target attribute types are possible: <ul> <li> regular: only regular attributes are used as input variables for learning tasks </li> <li> id: the id attribute for the example set </li> <li> label: target attribute for learning </li> <li> prediction: predicted attribute, i.e. the predictions of a learning scheme </li> <li> cluster: indicates the membership to a cluster </li> <li> weight: indicates the weight of the example </li> <li> batch: indicates the membership to an example batch </li> </ul> Users can also define own attribute types by simply using the desired name. Please be aware that roles have to be unique! Assigning a non regular role the second time will cause the first attribute to be dropped from the example set. If you want to keep this attribute, you have to change it's role first.

RM_Set_Role_prediction(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute,?AttName), representsAttribute(?Column,?Attribute), simpleParameter_target_role(?this, "prediction") -> copy(?Dout,?Din,{hasAttribute(?Din, ?Attribute)}), predictedAttribute(?Dout, ?Attribute), produces(?this,?Dout)

containsColumn(?Din, ?Column), hasAttribute(?Din, ?Attribute), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName) -> RM_Set_Role_prediction(?this), simpleParameter_attribute(?this, ?AttName), simpleParameter_include_special_attributes(?this, 1), simpleParameter_target_role(?this, "prediction")

set_role

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey name hasDefalutValue comment The name of the attribute whose role should be changed. hasParameterKey target_role hasDefalutValue regular comment The target role of the attribute (only changed if parameter change_attribute_type is true). 1 1 This operator can be used to change the role of an attribute of the input ExampleSet. If you want to change the attribute name you should use the <a href="rm://opdoc/rename">Rename</a> operator. The target role indicates if the attribute is a regular attribute (used by learning operators) or a special attribute (e.g. a label or id attribute). The following target attribute types are possible: <ul> <li> regular: only regular attributes are used as input variables for learning tasks </li> <li> id: the id attribute for the example set </li> <li> label: target attribute for learning </li> <li> prediction: predicted attribute, i.e. the predictions of a learning scheme </li> <li> cluster: indicates the membership to a cluster </li> <li> weight: indicates the weight of the example </li> <li> batch: indicates the membership to an example batch </li> </ul> Users can also define own attribute types by simply using the desired name. Please be aware that roles have to be unique! Assigning a non regular role the second time will cause the first attribute to be dropped from the example set. If you want to keep this attribute, you have to change it's role first.

set_role

RM_Set_Role_weight(?this), uses(?this, ?Din), simpleParameter_attribute(?this, ?AttName), attributeName(?Attribute,?AttName), representsAttribute(?Column,?Attribute), simpleParameter_target_role(?this, "weight") -> copy(?Dout,?Din,{hasAttribute(?Din, ?Attribute)}), weightingAttribute(?Dout, ?Attribute), produces(?this,?Dout)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

containsColumn(?Din, ?Column), hasAttribute(?Din, ?Attribute), representsAttribute(?Column, ?Attribute), attributeName(?Attribute,?AttName) -> RM_Set_Role_weight(?this), simpleParameter_attribute(?this, ?AttName), simpleParameter_include_special_attributes(?this, 1), simpleParameter_target_role(?this, "weight")

1 1

single_rule_induction

Parameters: hasParameterKey max_depth hasDefalutValue 2 comment An upper bound for the number of literals. hasParameterKey utility_function hasDefalutValue weighted relative accuracy comment The function to be optimized by the rule. hasParameterKey max_cache hasDefalutValue 10000 comment Bounds the number of rules considered per depth to avoid high memory consumption, but leads to incomplete search. hasParameterKey relative_to_predictions hasDefalutValue false comment Searches for rules with a maximum difference to the predited label. This operator returns the best rule regarding WRAcc using exhaustive search. Features like the incorporation of other metrics and the search for more than a single rule are prepared. The search strategy is BFS, with save pruning whenever applicable. This operator can easily be extended to support other search strategies.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_Single_Rule_Induction(?this) ->

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_Single_Rule_Induction(?this)

1 1 This operator concentrates on one single attribute and determines the best splitting terms for minimizing the training error. The result will be a single rule containing all these terms.

RM_Single_Rule_Induction_Single_Attribute(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters:

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn or CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_Single_Rule_Induction_Single_Attribute(?this)

single_rule_induction_single_attribute

1 1

RM_Subgroup_Discovery(?this) ->

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (CategorialColumn)) and MissingValueFreeDataTable](?D) -> RM_Subgroup_Discovery(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey mode hasDefalutValue k best rules comment Discovery mode. hasParameterKey utility_function hasDefalutValue WRAcc comment Utility function. hasParameterKey min_utility hasDefalutValue 0.0 comment Minimum quality which has to be reached. hasParameterKey k_best_rules hasDefalutValue 10 comment Report the k best rules. hasParameterKey rule_generation hasDefalutValue both comment Determines which rules are generated. hasParameterKey max_depth hasDefalutValue 5 comment Maximum depth of BFS. hasParameterKey min_coverage hasDefalutValue 0.0 comment Only consider rules which exceed the given coverage threshold. hasParameterKey max_cache hasDefalutValue -1 comment Bounds the number of rules which are evaluated (only the most supported rules are used). This operator discovers subgroups (or induces a rule set, respectively) by generating hypotheses exhaustively. Generation is done by stepwise refining the empty hypothesis (which contains no literals). The loop for this task hence iterates over the depth of the search space, i.e. the number of literals of the generated hypotheses. The maximum depth of the search can be specified. Furthermore the search space can be pruned by specifying a minimum coverage of the hypothesis or by using only a given amount of hypotheses which have the highest coverage. From the hypotheses, rules are derived according to the users preference. The operator allows the derivation of positive rules (Y+) and negative rules (Y-) separately or the combination by deriving both rules or only the one which is the most probable due to the examples covered by the hypothesis (hence: the actual prediction for that subset). All generated rules are evaluated on the example set by a user specified utility function and stored in the final rule set if they (1) exceed a minimum utility threshold or (2) are among the k best rules. The desired behavior can be specified as well.

subgroup_discovery

1 1 1 Parameters: hasParameterKey add_cluster_attribute hasDefalutValue true comment Indicates if a cluster id is generated as new special attribute. hasParameterKey add_as_label hasDefalutValue false comment Should the cluster values be added as label. hasParameterKey remove_unlabeled hasDefalutValue false comment Delete the unlabeled examples. hasParameterKey min_pts hasDefalutValue 2 comment The minimal number of points in each cluster. hasParameterKey kernel_type hasDefalutValue radial comment The SVM kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The SVM kernel parameter gamma (radial). hasParameterKey kernel_degree hasDefalutValue 2 comment The SVM kernel parameter degree (polynomial). hasParameterKey kernel_a hasDefalutValue 1.0 comment The SVM kernel parameter a (neural). hasParameterKey kernel_b hasDefalutValue 0.0 comment The SVM kernel parameter b (neural). hasParameterKey kernel_cache hasDefalutValue 200 comment Size of the cache for kernel evaluations im MB hasParameterKey convergence_epsilon hasDefalutValue 0.0010 comment Precision on the KKT conditions hasParameterKey max_iterations hasDefalutValue 100000 comment Stop after this many iterations hasParameterKey p hasDefalutValue 0.0 comment The fraction of allowed outliers. hasParameterKey r hasDefalutValue -1.0 comment Use this radius instead of the calculated one (-1 for calculated radius). hasParameterKey number_sample_points hasDefalutValue 20 comment The number of virtual sample points to check for neighborship. An implementation of Support Vector Clustering based on {@rapidminer.cite BenHur/etal/2001a}. This operator will create a cluster attribute if not present yet.

[DataTable and (inputColumn only (ScalarColumn))](?D) -> RM_Support_Vector_Clustering(?this)

support_vector_clustering

RM_Support_Vector_Clustering(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey kernel_type hasDefalutValue dot comment The SVM kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The SVM kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The SVM kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The SVM kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The SVM kernel parameter sigma3. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The SVM kernel parameter shift. hasParameterKey kernel_degree hasDefalutValue 2.0 comment The SVM kernel parameter degree. hasParameterKey kernel_a hasDefalutValue 1.0 comment The SVM kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The SVM kernel parameter b. hasParameterKey kernel_cache hasDefalutValue 200 comment Size of the cache for kernel evaluations im MB hasParameterKey C hasDefalutValue 0.0 comment The SVM complexity constant. Use -1 for different C values for positive and negative. hasParameterKey convergence_epsilon hasDefalutValue 0.0010 comment Precision on the KKT conditions hasParameterKey max_iterations hasDefalutValue 100000 comment Stop after this many iterations hasParameterKey scale hasDefalutValue true comment Scale the example values and store the scaling parameters for test set. hasParameterKey calculate_weights hasDefalutValue true comment Indicates if attribute weights should be returned. hasParameterKey return_optimization_performance hasDefalutValue true comment Indicates if final optimization fitness should be returned as performance. hasParameterKey L_pos hasDefalutValue 1.0 comment A factor for the SVM complexity constant for positive examples hasParameterKey L_neg hasDefalutValue 1.0 comment A factor for the SVM complexity constant for negative examples hasParameterKey epsilon hasDefalutValue 0.0 comment Insensitivity constant. No loss if prediction lies this close to true value hasParameterKey epsilon_plus hasDefalutValue 0.0 comment Epsilon for positive deviation only hasParameterKey epsilon_minus hasDefalutValue 0.0 comment Epsilon for negative deviation only hasParameterKey balance_cost hasDefalutValue false comment Adapts Cpos and Cneg to the relative size of the classes hasParameterKey quadratic_loss_pos hasDefalutValue false comment Use quadratic loss for positive deviation hasParameterKey quadratic_loss_neg hasDefalutValue false comment Use quadratic loss for negative deviation hasParameterKey estimate_performance hasDefalutValue false comment Indicates if this learner should also return a performance estimation. This learner uses the Java implementation of the support vector machine mySVM by Stefan Rüping. This learning method can be used for both regression and classification and provides a fast algorithm and good results for many learning tasks.

RM_Support_Vector_Machine(?this) ->

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Support_Vector_Machine(?this)

support_vector_machine

1 1 This is a SVM implementation using an evolutionary algorithm (ES) to solve the dual optimization problem of a SVM. It turns out that on many datasets this simple implementation is as fast and accurate as the usual SVM implementations. In addition, it is also capable of learning with Kernels which are not positive semi-definite and can also be used for multi-objective learning which makes the selection of C unecessary before learning. Mierswa, Ingo. Evolutionary Learning with Kernels: A Generic Solution for Large Margin Problems. In Proc. of the Genetic and Evolutionary Computation Conference (GECCO 2006), 2006.

RM_Support_Vector_Machine_Evolutionary(?this) ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b. hasParameterKey C hasDefalutValue 0.0 comment The SVM complexity constant (0: calculates probably good value). hasParameterKey epsilon hasDefalutValue 0.1 comment The width of the regression tube loss function of the regression SVM hasParameterKey start_population_type hasDefalutValue random comment The type of start population initialization. hasParameterKey max_generations hasDefalutValue 10000 comment Stop after this many evaluations hasParameterKey generations_without_improval hasDefalutValue 30 comment Stop after this number of generations without improvement (-1: optimize until max_iterations). hasParameterKey population_size hasDefalutValue 1 comment The population size (-1: number of examples) hasParameterKey tournament_fraction hasDefalutValue 0.75 comment The fraction of the population used for tournament selection. hasParameterKey keep_best hasDefalutValue true comment Indicates if the best individual should survive (elititst selection). hasParameterKey mutation_type hasDefalutValue gaussian_mutation comment The type of the mutation operator. hasParameterKey selection_type hasDefalutValue tournament comment The type of the selection operator. hasParameterKey crossover_prob hasDefalutValue 1.0 comment The probability for crossovers. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed hasParameterKey hold_out_set_ratio hasDefalutValue 0.0 comment Uses this amount as a hold out set to estimate generalization error after learning (currently only used for multi-objective classification). hasParameterKey show_convergence_plot hasDefalutValue false comment Indicates if a dialog with a convergence plot should be drawn. hasParameterKey show_population_plot hasDefalutValue false comment Indicates if the population plot in case of the non-dominated sorting should be shown. hasParameterKey return_optimization_performance hasDefalutValue false comment Indicates if final optimization fitness should be returned as performance.

support_vector_machine_evolutionary

[DataTable and (targetColumn exactly 1 (ScalarColumn or BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Support_Vector_Machine_Evolutionary(?this)

1 1 Applies the <a href="http://www.csie.ntu.edu.tw/~cjlin/libsvm">libsvm</a> learner by Chih-Chung Chang and Chih-Jen Lin. The SVM is a powerful method for both classification and regression. This operator supports the SVM types <code>C-SVC</code> and <code>nu-SVC</code> for classification tasks as well as <code>epsilon-SVR</code> and <code>nu-SVR</code> for regression tasks. Additionally <code>one-class</code> gives the possibility to learn from just one class of examples and lateron test if new examples match the knowing ones. In contrast to other SVM learners, the libsvm supports also internal multiclass learning and probability estimation based on Platt scaling for proper confidence values after applying the learned model on a classification data set.

RM_Support_Vector_Machine_LibSVM_C_SVC(?this), simpleParameter_svm_type(?this, "C-SVC") ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey svm_type hasDefalutValue C-SVC comment SVM for classification (C-SVC, nu-SVC), regression (epsilon-SVR, nu-SVR) and distribution estimation (one-class) hasParameterKey kernel_type hasDefalutValue rbf comment The type of the kernel functions hasParameterKey degree hasDefalutValue 3 comment The degree for a polynomial kernel function. hasParameterKey gamma hasDefalutValue 0.0 comment The parameter gamma for polynomial, rbf, and sigmoid kernel functions (0 means 1/#examples). hasParameterKey coef0 hasDefalutValue 0.0 comment The parameter coef0 for polynomial and sigmoid kernel functions. hasParameterKey C hasDefalutValue 0.0 comment The cost parameter C for c_svc, epsilon_svr, and nu_svr. hasParameterKey nu hasDefalutValue 0.5 comment The parameter nu for nu_svc, one_class, and nu_svr. hasParameterKey cache_size hasDefalutValue 80 comment Cache size in Megabyte. hasParameterKey epsilon hasDefalutValue 0.0010 comment Tolerance of termination criterion. hasParameterKey p hasDefalutValue 0.1 comment Tolerance of loss function of epsilon-SVR. hasParameterKey class_weights hasDefalutValue comment The weights w for all classes (first column: class name, second column: weight), i.e. set the parameters C of each class w * C (empty: using 1 for all classes where the weight was not defined). hasParameterKey shrinking hasDefalutValue true comment Whether to use the shrinking heuristics. hasParameterKey calculate_confidences hasDefalutValue false comment Indicates if proper confidence values should be calculated. hasParameterKey confidence_for_multiclass hasDefalutValue true comment Indicates if the class with the highest confidence should be selected in the multiclass setting. Uses binary majority vote over all 1-vs-1 classifiers otherwise (selected class must not be the one with highest confidence in that case).

[DataTable and (targetColumn exactly 1 (CategorialColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Support_Vector_Machine_LibSVM_C_SVC(?this), simpleParameter_svm_type(?this, "C-SVC")

support_vector_machine_libsvm

1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

support_vector_machine_libsvm

[DataTable and (targetColumn exactly 1 (ScalarColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Support_Vector_Machine_LibSVM_epsilon_SVR(?this), simpleParameter_svm_type(?this, "epsilon-SVR")

RM_Support_Vector_Machine_LibSVM_epsilon_SVR(?this), simpleParameter_svm_type(?this, "epsilon-SVR") ->

Applies the <a href="http://www.csie.ntu.edu.tw/~cjlin/libsvm">libsvm</a> learner by Chih-Chung Chang and Chih-Jen Lin. The SVM is a powerful method for both classification and regression. This operator supports the SVM types <code>C-SVC</code> and <code>nu-SVC</code> for classification tasks as well as <code>epsilon-SVR</code> and <code>nu-SVR</code> for regression tasks. Additionally <code>one-class</code> gives the possibility to learn from just one class of examples and lateron test if new examples match the knowing ones. In contrast to other SVM learners, the libsvm supports also internal multiclass learning and probability estimation based on Platt scaling for proper confidence values after applying the learned model on a classification data set. 1 1 Parameters: hasParameterKey show_convergence_plot hasDefalutValue false comment Indicates if a dialog with a convergence plot should be drawn. hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b. hasParameterKey C hasDefalutValue 0.0 comment The SVM complexity constant (0: calculates probably good value). hasParameterKey max_evaluations hasDefalutValue 500 comment Stop after this many evaluations hasParameterKey generations_without_improval hasDefalutValue 10 comment Stop after this number of generations without improvement (-1: optimize until max_iterations). hasParameterKey population_size hasDefalutValue 10 comment The population size (-1: number of examples) hasParameterKey inertia_weight hasDefalutValue 0.1 comment The (initial) weight for the old weighting. hasParameterKey local_best_weight hasDefalutValue 1.0 comment The weight for the individual's best position during run. hasParameterKey global_best_weight hasDefalutValue 1.0 comment The weight for the population's best position during run. hasParameterKey dynamic_inertia_weight hasDefalutValue true comment If set to true the inertia weight is improved during run. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed This is a SVM implementation using a particle swarm optimization (PSO) approach to solve the dual optimization problem of a SVM. It turns out that on many datasets this simple implementation is as fast and accurate as the usual SVM implementations.

RM_Support_Vector_Machine_PSO(?this) ->

support_vector_machine_pso

[DataTable and (targetColumn exactly 1 (BinaryColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Support_Vector_Machine_PSO(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 This operator performs a vector linear regression. It regresses all regular attributes upon a vector of labels. The attributes forming the vector have to be marked as special, the special role names of all label attributes have to start with <code>label</code>. TODO: Adapt meta data of model, but needs change of complete construction...

[DataTable and (targetColumn exactly 1 (ScalarColumn)) and (inputColumn only (ScalarColumn)) and MissingValueFreeDataTable](?D) -> RM_Vector_Linear_Regression(?this)

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

Parameters: hasParameterKey use_bias hasDefalutValue true comment Indicates if an intercept value should be calculated. hasParameterKey ridge hasDefalutValue 1.0E-8 comment The ridge parameter.

vector_linear_regression

RM_Vector_Linear_Regression(?this) ->

1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

amountOfRows(?Din, ?NumRows) -> RM_WeightedBootstrapping(?this)

RM_WeightedBootstrapping(?this), uses(?this, ?Din) -> copy(?Dout, ?Din, {amountOfRows(?Din, ?_)}), produces(?this, ?Dout)

Parameters: hasParameterKey sample_ratio hasDefalutValue 1.0 comment This ratio determines the size of the new example set. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed

weightedbootstrapping

This operator constructs a bootstrapped sample from the given example set which must provide a weight attribute. If no weight attribute was provided this operator will stop the process with an error message. See the operator Bootstrapping for more information. 1 1 1

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator represents an implementation of k-means. This operator will create a cluster attribute if not present yet.

k_means

[DataTable and (inputColumn only (ScalarColumn))](?D) -> RM_k_Means(?this)

RM_k_Means(?this) ->

Parameters: hasParameterKey add_cluster_attribute hasDefalutValue true comment Indicates if a cluster id is generated as new special attribute. hasParameterKey add_as_label hasDefalutValue false comment Should the cluster values be added as label. hasParameterKey remove_unlabeled hasDefalutValue false comment Delete the unlabeled examples. hasParameterKey k hasDefalutValue 2 comment The number of clusters which should be detected. hasParameterKey max_runs hasDefalutValue 10 comment The maximal number of runs of k-Means with random initialization that are performed. hasParameterKey max_optimization_steps hasDefalutValue 100 comment The maximal number of iterations performed for one run of k-Means. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed 1 1 1 Parameters: hasParameterKey add_cluster_attribute hasDefalutValue true comment Indicates if a cluster id is generated as new special attribute. hasParameterKey add_as_label hasDefalutValue false comment Should the cluster values be added as label. hasParameterKey remove_unlabeled hasDefalutValue false comment Delete the unlabeled examples. hasParameterKey k hasDefalutValue 2 comment The number of clusters which should be detected. hasParameterKey max_runs hasDefalutValue 10 comment The maximal number of runs of k-Means with random initialization that are performed. hasParameterKey max_optimization_steps hasDefalutValue 100 comment The maximal number of iterations performed for one run of k-Means. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed hasParameterKey measure_types hasDefalutValue MixedMeasures comment The measure type hasParameterKey mixed_measure hasDefalutValue MixedEuclideanDistance comment Select measure hasParameterKey nominal_measure hasDefalutValue NominalDistance comment Select measure hasParameterKey numerical_measure hasDefalutValue EuclideanDistance comment Select measure hasParameterKey divergence hasDefalutValue GeneralizedIDivergence comment Select divergence hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b.

[DataTable and (inputColumn only (NormalizedScalarColumn))](?D) -> RM_k_Medoids_BregmanDivergences(?this), simpleParameter_measure_types(?this, "BregmanDivergences")

RM_k_Medoids_BregmanDivergences(?this), simpleParameter_measure_types(?this, "BregmanDivergences") ->

k_medoids

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

This operator represents an implementation of k-medoids. This operator will create a cluster attribute if not present yet. 1 1 1 This operator represents an implementation of k-medoids. This operator will create a cluster attribute if not present yet.

RM_k_Medoids_MixedMeasures(?this), simpleParameter_measure_types(?this, "MixedMeasures") ->

[DataTable and (inputColumn only (NormalizedScalarColumn or CategorialColumn))](?D) -> RM_k_Medoids_MixedMeasures(?this), simpleParameter_measure_types(?this, "MixedMeasures")

k_medoids

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

1 1 1 This operator represents an implementation of k-medoids. This operator will create a cluster attribute if not present yet.

RM_k_Medoids_NominalMeasures(?this), simpleParameter_measure_types(?this, "NominalMeasures") ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

k_medoids

[DataTable and (inputColumn only (CategorialColumn))](?D) -> RM_k_Medoids_NominalMeasures(?this), simpleParameter_measure_types(?this, "NominalMeasures")

1 1 1

RM_k_Medoids_NumericalMeasures(?this), simpleParameter_measure_types(?this, "NumericalMeasures") ->

k_medoids

This operator represents an implementation of k-medoids. This operator will create a cluster attribute if not present yet. Parameters: hasParameterKey add_cluster_attribute hasDefalutValue true comment Indicates if a cluster id is generated as new special attribute. hasParameterKey add_as_label hasDefalutValue false comment Should the cluster values be added as label. hasParameterKey remove_unlabeled hasDefalutValue false comment Delete the unlabeled examples. hasParameterKey k hasDefalutValue 2 comment The number of clusters which should be detected. hasParameterKey max_runs hasDefalutValue 10 comment The maximal number of runs of k-Means with random initialization that are performed. hasParameterKey max_optimization_steps hasDefalutValue 100 comment The maximal number of iterations performed for one run of k-Means. hasParameterKey use_local_random_seed hasDefalutValue false comment Indicates if a local random seed should be used. hasParameterKey local_random_seed hasDefalutValue 1992 comment Specifies the local random seed hasParameterKey measure_types hasDefalutValue MixedMeasures comment The measure type hasParameterKey mixed_measure hasDefalutValue MixedEuclideanDistance comment Select measure hasParameterKey nominal_measure hasDefalutValue NominalDistance comment Select measure hasParameterKey numerical_measure hasDefalutValue EuclideanDistance comment Select measure hasParameterKey divergence hasDefalutValue GeneralizedIDivergence comment Select divergence hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (inputColumn only (NormalizedScalarColumn))](?D) -> RM_k_Medoids_NumericalMeasures(?this), simpleParameter_measure_types(?this, "NumericalMeasures")

1 1

k_nn

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (NormalizedScalarColumn))](?D) -> RM_k_NN_BregmanDivergences(?this), simpleParameter_measure_types(?this, "BregmanDivergences")

A k nearest neighbor implementation.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_k_NN_BregmanDivergences(?this), simpleParameter_measure_types(?this, "BregmanDivergences") ->

Parameters: hasParameterKey k hasDefalutValue 1 comment The used number of nearest neighbors. hasParameterKey weighted_vote hasDefalutValue false comment Indicates if the votes should be weighted by similarity. hasParameterKey measure_types hasDefalutValue MixedMeasures comment The measure type hasParameterKey mixed_measure hasDefalutValue MixedEuclideanDistance comment Select measure hasParameterKey nominal_measure hasDefalutValue NominalDistance comment Select measure hasParameterKey numerical_measure hasDefalutValue EuclideanDistance comment Select measure hasParameterKey divergence hasDefalutValue GeneralizedIDivergence comment Select divergence hasParameterKey kernel_type hasDefalutValue radial comment The kernel type hasParameterKey kernel_gamma hasDefalutValue 1.0 comment The kernel parameter gamma. hasParameterKey kernel_sigma1 hasDefalutValue 1.0 comment The kernel parameter sigma1. hasParameterKey kernel_sigma2 hasDefalutValue 0.0 comment The kernel parameter sigma2. hasParameterKey kernel_sigma3 hasDefalutValue 2.0 comment The kernel parameter sigma3. hasParameterKey kernel_degree hasDefalutValue 3.0 comment The kernel parameter degree. hasParameterKey kernel_shift hasDefalutValue 1.0 comment The kernel parameter shift. hasParameterKey kernel_a hasDefalutValue 1.0 comment The kernel parameter a. hasParameterKey kernel_b hasDefalutValue 0.0 comment The kernel parameter b. 1 1

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (NormalizedScalarColumn or CategorialColumn))](?D) -> RM_k_NN_MixedMeasures(?this), simpleParameter_measure_types(?this, "MixedMeasures")

RM_k_NN_MixedMeasures(?this), simpleParameter_measure_types(?this, "MixedMeasures") ->

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

k_nn

A k nearest neighbor implementation. 1 1

k_nn

A k nearest neighbor implementation.

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (CategorialColumn))](?D) -> RM_k_NN_NominalMeasures(?this), simpleParameter_measure_types(?this, "NominalMeasures")

RM_k_NN_NominalMeasures(?this), simpleParameter_measure_types(?this, "NominalMeasures") ->

1 1

[DataTable and (targetColumn exactly 1 (ScalarColumn or CategorialColumn)) and (inputColumn only (NormalizedScalarColumn))](?D) -> RM_k_NN_NumericalMeasures(?this), simpleParameter_measure_types(?this, "NumericalMeasures")

A k nearest neighbor implementation.

k_nn

http://rapid-i.dyndns.org:8080/e-LICO-Web/ExecutorService?wsdl

RM_k_NN_NumericalMeasures(?this), simpleParameter_measure_types(?this, "NumericalMeasures") ->

[DataTable and (inputAttribute min 1 Attribute) and (inputColumn only (DataColumn and columnHasType only (Scalar or Categorial))) ](?D) -> new(?this), UnsupervisedLearner(?this), uses(?this,?D)