Module es.upm.fi.cig.multictbnc

Package es.upm.fi.cig.multictbnc.experiments.implementationsexperiments.featurestreamexperiments

Class FeatureStreamExperimentFactory

java.lang.Object

es.upm.fi.cig.multictbnc.experiments.implementationsexperiments.featurestreamexperiments.FeatureStreamExperimentFactory

public class FeatureStreamExperimentFactory
extends Object

Factory class for creating specific types of feature stream experiments.

Constructor Summary

Constructors

Constructor	Description
FeatureStreamExperimentFactory()

Method Summary

Modifier and Type	Method	Description
static FeatureStreamImplementationExperiment	getFeatureStreamImplementation(String typeFeatureStreamProcessing, String pathFeatureStream, String nameTimeVariable, List<String> nameClassVariables, DatasetReader initialDatasetReader, DatasetReader testDatasetReader, CTBNParameterLearningAlgorithm ctbnPLA, int maxSeparatingSizeFSS, int experimentNumber, int numFeatures, boolean areModelsValidated, double sigTimeTransitionHyp, double sigStateToStateTransitionHyp)	Creates an instance of a feature stream experiment based on the specified type of processing.

Methods inherited from class java.lang.Object

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

Constructor Details

FeatureStreamExperimentFactory

public FeatureStreamExperimentFactory()

Method Details

getFeatureStreamImplementation

public static FeatureStreamImplementationExperiment getFeatureStreamImplementation(String typeFeatureStreamProcessing,
 String pathFeatureStream,
 String nameTimeVariable,
 List<String> nameClassVariables,
 DatasetReader initialDatasetReader,
 DatasetReader testDatasetReader,
 CTBNParameterLearningAlgorithm ctbnPLA,
 int maxSeparatingSizeFSS,
 int experimentNumber,
 int numFeatures,
 boolean areModelsValidated,
 double sigTimeTransitionHyp,
 double sigStateToStateTransitionHyp)
                                                                            throws UnreadDatasetException,
IOException

Creates an instance of a feature stream experiment based on the specified type of processing. Different types of processing methods like 'Retraining(FSS)', 'Batch(FSS)', 'Batch', and 'Retraining' are supported, each corresponding to a different implementation strategy for handling feature streams.

Parameters:

typeFeatureStreamProcessing - type of feature stream processing method

pathFeatureStream - path to the feature stream

nameTimeVariable - name of the time variable

nameClassVariables - list of names of class variables

initialDatasetReader - reader for the initial dataset

testDatasetReader - reader for the test dataset

ctbnPLA - CTBN parameter learning algorithm

maxSeparatingSizeFSS - maximum separating size for feature subset selection

experimentNumber - identifier of the experiment

numFeatures - number of features in the dataset

areModelsValidated - boolean flag indicating if models should be validated

sigTimeTransitionHyp - significance level for time transition hypothesis

sigStateToStateTransitionHyp - significance level for state-to-state transition hypothesis

Returns:

instance of FeatureStreamImplementationExperiment

Throws:

UnreadDatasetException - if the dataset cannot be read

IOException - if an error occurs in file handling