netkit.classifiers.relational
Class NetworkMetaClassifier

java.lang.Object
  netkit.classifiers.ClassifierImp
      netkit.classifiers.relational.NetworkClassifierImp
          netkit.classifiers.relational.NetworkMetaClassifier

All Implemented Interfaces:

Classifier, NetworkClassifier, Configurable

Direct Known Subclasses:

MetaMultiplicative

public abstract class NetworkMetaClassifier
extends NetworkClassifierImp

Abstract class for combining multiple relational and non-relational classifiers. It configures itsel with the list of (non-)relational classifiers to use and induces them individually. How to combine them is left to subclasses. It uses the Networklearning.LC_PREFIX and Networklearning.RC_PREFIX properties from the configuration details specified in the lclassifier.properties and rclassifier.properties files (in addition to any properties used by the superclass).

Properties

• NetworkLearning.LC_PREFIX: what non-relational classifiers to use. default=naivebayes (see lclassifier.properties)
• NetworkLearning.RC_PREFIX: what relational classifiers to use. default=naivebayes (see rclassifier.properties)

Author:

Sofus A. Macskassy (sofmac@gmail.com)

See Also:

NetworkLearning.LC_PREFIX, NetworkLearning.RC_PREFIX

Nested Class Summary

Nested classes/interfaces inherited from class netkit.classifiers.relational.NetworkClassifierImp

NetworkClassifierImp.Aggregation

Field Summary

protected java.util.ArrayList<Classifier>	lclassifiers The list of non-relational classifiers to use
protected java.util.ArrayList<NetworkClassifier>	rclassifiers The list of relational classifiers to use

Fields inherited from class netkit.classifiers.relational.NetworkClassifierImp

aggFactory, aggregation, aggregators, aggTypes, dynamicAggregators, prior

Fields inherited from class netkit.classifiers.ClassifierImp

attribute, classPrior, clsIdx, graph, keyIndex, logger, nodeType, right, tmpVector, useIntrinsic, vectorClsIdx

Constructor Summary

NetworkMetaClassifier()

Method Summary

void	configure(Configuration config) Configure this classifier by getting the Weka classifier object using the classifier and options properties in addition to anything used by the superclass.
protected java.lang.String	getClassifierNames() Returns the list of classifier names in the format: "RC[relational_classifiers] LC[nonrelational_classifiers]" where the list of classifiers is comma-separated and appear exactly is it was in the configuration object that was used to configure this classifier.
Configuration	getDefaultConfiguration() Get the detault configuration of using a naive Bayes classifier both as the single non-relational and the single relational classifier..
void	induceModel(Graph graph, DataSplit split) This separately induces all the non-relational and relational classifiers in addition to any setup the super-class needs to do.

Methods inherited from class netkit.classifiers.relational.NetworkClassifierImp

classify, doEstimate, estimate, estimate, estimate, estimate, generateAggregators, getAttributeNames, includeClassAttribute, initializeRun, makeVector

Methods inherited from class netkit.classifiers.ClassifierImp

addListener, classify, classify, clearListeners, estimate, estimate, getLogger, getNofifyListeners, notifyListeners, notifyListeners, removeListener, reset, setNofityListeners

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface netkit.classifiers.Classifier

addListener, classify, classify, clearListeners, estimate, estimate, getDescription, getLogger, getName, getNofifyListeners, getShortName, notifyListeners, notifyListeners, removeListener, reset, setNofityListeners

Field Detail

rclassifiers

protected java.util.ArrayList<NetworkClassifier> rclassifiers

The list of relational classifiers to use

lclassifiers

protected java.util.ArrayList<Classifier> lclassifiers

The list of non-relational classifiers to use

Constructor Detail

NetworkMetaClassifier

public NetworkMetaClassifier()

Method Detail

getDefaultConfiguration

public Configuration getDefaultConfiguration()

Get the detault configuration of using a naive Bayes classifier both as the single non-relational and the single relational classifier.. This is in addition to any defaults set by the superclass.

Specified by:

getDefaultConfiguration in interface Configurable

Overrides:

getDefaultConfiguration in class NetworkClassifierImp

Returns:

a Configuration object

See Also:

NetworkClassifierImp.getDefaultConfiguration()

configure

public void configure(Configuration config)

Configure this classifier by getting the Weka classifier object using the classifier and options properties in addition to anything used by the superclass. The classifier property defines the weka classifier (this name should resolve to a real weka class name in the weka.properties file). The options property defines the options string to pass to the weka classifier.

By default, this will use a naive Bayes classifier both as the single non-relational and the single relational classifier.

Specified by:

configure in interface Configurable

Overrides:

configure in class NetworkClassifierImp

Parameters:

config - The configuration object used to configure this classifier

See Also:

NetworkClassifierImp.configure(netkit.util.Configuration)

induceModel

public void induceModel(Graph graph,
                        DataSplit split)

This separately induces all the non-relational and relational classifiers in addition to any setup the super-class needs to do.

Specified by:

induceModel in interface Classifier

Overrides:

induceModel in class NetworkClassifierImp

Parameters:

graph - The graph over which a model is induced

split - The datasplit that specifies which nodes have their class labels known

See Also:

NetworkClassifierImp.induceModel(netkit.graph.Graph, netkit.classifiers.DataSplit)

getClassifierNames

protected java.lang.String getClassifierNames()

Returns the list of classifier names in the format: "RC[relational_classifiers] LC[nonrelational_classifiers]" where the list of classifiers is comma-separated and appear exactly is it was in the configuration object that was used to configure this classifier.

Returns:

the list of classifier names from the configuration object when configure was called

See Also:

configure(netkit.util.Configuration)