Package oracle.dmt.odm.settings.function

This package contains Java classes representing mining function settings.

See:
Description

Class Summary

AssociationRulesFunctionSettings	An instance of AssociationRulesFunctionSettings describes settings for an association rules model.
AttributeImportanceFunctionSettings	An instance of AttributeImportanceFunctionSettings describes the settings for Attribute Importance function that captures the high-level specification for building an attribute importance model.
ClassificationFunctionSettings	An instance of ClassificationFunctionSettings describes the settings necessary to build a classification model.
ClusteringFunctionSettings	An instance of ClusteringFunctionSettings holds metadata for required settings common to all clustering algorithms.
DataUsageEntry	An instance of DataUsageEntry specifies how to use a particular mining attribute in the LogicalDataSpecification of a given MiningFunctionSettings
DataUsageSpecification	An instance of DataUsageSpecification is used to specify how the attributes in a LogicalDataSpecification instance are used for building a mining model.
FeatureExtractionFunctionSettings	FeatureExtractionFunctionSettings describes settings for Feature Extraction functions.
MiningFunctionSettings	The abstract class MiningFunctionSettings (MFS) captures the high level specification input for building a data mining model.
RegressionFunctionSettings	An instance of RegressionFunctionSettings describes the settings necessary to build a Regression (regression) model by using a Support Vector Machine mining algorithm.
SupervisedFunctionSettings	SupervisedFunctionSettings describes settings for supervised learning functions.

Package oracle.dmt.odm.settings.function Description

This package contains Java classes representing mining function settings. Mining function settings are named objects that application programmers construct and provide to Oracle Data Mining to build a model. It consists of a LogicalDataSpecification and maps the physical input data to the logical form understood by ODM, DataUsageSpecification that specifies how each mining attribute is to be used for the build operation, the type of function to be executed, and an optional algorithm settings. An instance of mining function settings must be stored in the DMS before it can be used to build a model.

A mining function settings object captures the high level specification input for building a data mining model. In ODM, the functions are divided into the key areas: classification, approximation, association rules, and clustering. The intent of function settings is to allow a user to specify the type of result desired without having to specify a particular algorithm. Although mining function settings allow for the specification of algorithm, if this is omitted, the underlying DMS is responsible for selecting the algorithm based on basic user-provided parameters.

ODM supports the following types of mining function settings:

• AssociationRulesFunctionSettings
• AttributeImportanceFunctionSettings
• ClusteringFunctionSettings
• SupervisedFunctionSettings: A place holder for all supervised function settings
  • ClassificationFunctionSettings
  • ApproximationFunctionSettings
• FeatureExtractionFunctionSettings