emolib.classifier.machinelearning
Class Logistic

java.lang.Object
  emolib.util.proc.TextDataProcessor
      emolib.classifier.Classifier
          emolib.classifier.machinelearning.Logistic

All Implemented Interfaces:

Configurable, DataProcessor

public class Logistic
extends Classifier

The Logistic class is a Multinomial Logistic Regression (MLR) classifier.

It is a probabilistic discriminative approach that fits a set of exponential functions via the Maximum A Posteriori estimation. MLR obeys the maximum entropy principle, therefore it does not make any further assumption beyond what is directly observed in the training data. Moreover, it makes no assumptions about the relationships among the features, and so might potentially be more effective when conditional independence assumptions are not met, also overcoming the sparseness problem.

The core implementation of this logistic regression is based on LingPipe. The same term weighting schemes as the ones used in the ARN-R are considered.

Author:

Alexandre Trilla (atrilla@salle.url.edu)

See Also:

ARNReduced

Nested Class Summary

class

Logistic.VectorStatistics Inner class to compute vector statistics.

Constructor Summary

Logistic()
Main constructor of this logistic regression classifier.

Method Summary

java.lang.String	getCategory(FeatureBox inputFeatures) The function that decides the most appropriate emotional category.
void	load(java.lang.String path) Generic function to load a previously saved classifier.
static void	main(java.lang.String[] args) Main method to test the LR classifier.
void	resetExamples() Method to reset the classifier and flush the training examples.
void	save(java.lang.String path) Generic method to save the fully fledged classifier into a given file path.
void	setChi2(boolean chi, int numF) Set the Chi square feature selection.
void	setCOF(boolean cof) Method to consider bigram frequencies.
void	setEmotionDims(boolean emodims) Method to consider emotion dimensions.
void	setIntercept(boolean intercept) Method to consider the intercept feature.
void	setMI(boolean mi, int numF) Set the Mutual Information feature selection.
void	setNegation(boolean neg) Method to consider negations.
void	setPOS(boolean pos) Method to consider POS tags.
void	setPrior(java.lang.String prior) Method to set the regression prior.
void	setStemming(boolean stems) Method to consider stems.
void	setSynonyms(boolean syns) Method to consider synonyms.
void	setTermWeighingMeasure(java.lang.String twm) Method to set the TW method.
void	setTF(boolean tf, int numF) Set the Term Frequency feature selection.
void	trainingProcedure() Training method based on the Stochastic Gradient Descent.

Methods inherited from class emolib.classifier.Classifier

applyClassification, getData, getListOfExampleCategories, getListOfExampleFeatures, initialize, inputTrainingExample, newProperties, register, train

Methods inherited from class emolib.util.proc.TextDataProcessor

flush, getName, getPredecessor, setPredecessor, toString

Methods inherited from class java.lang.Object

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

Constructor Detail

Logistic

public Logistic()

Main constructor of this logistic regression classifier. The logistic regression object is actually initialised when Logistic is trained.

Method Detail

getCategory

public java.lang.String getCategory(FeatureBox inputFeatures)

Description copied from class: Classifier

The function that decides the most appropriate emotional category. This is required for any classifier. The classifier in question has to previously run any training algorithm in order to provide the required prediction.

Specified by:

getCategory in class Classifier

Parameters:

inputFeatures - The input emotional features.

Returns:

The most appropriate emotional category.

trainingProcedure

public void trainingProcedure()

Training method based on the Stochastic Gradient Descent.

Specified by:

trainingProcedure in class Classifier

setMI

public void setMI(boolean mi,
                  int numF)

Set the Mutual Information feature selection.

Parameters:

mi - The Mutual Information flag.

numF - The number of relevant features desired.

setChi2

public void setChi2(boolean chi,
                    int numF)

Set the Chi square feature selection.

Parameters:

chi - The Chi2 flag.

numF - The number of relevant features desired.

setTF

public void setTF(boolean tf,
                  int numF)

Set the Term Frequency feature selection.

Parameters:

tf - The Term Frequency flag.

numF - The number of relevant features desired.

setTermWeighingMeasure

public void setTermWeighingMeasure(java.lang.String twm)

Method to set the TW method.

Parameters:

twm - The TW method.

setCOF

public void setCOF(boolean cof)

Method to consider bigram frequencies.

Parameters:

cof - The COF flag.

setPOS

public void setPOS(boolean pos)

Method to consider POS tags.

Parameters:

pos - The POS flag.

setStemming

public void setStemming(boolean stems)

Method to consider stems.

Parameters:

stems - The stemming flag.

setIntercept

public void setIntercept(boolean intercept)

Method to consider the intercept feature.

Parameters:

intercept - The intercept flag.

setSynonyms

public void setSynonyms(boolean syns)

Method to consider synonyms.

Parameters:

syns - The synonyms flag.

setEmotionDims

public void setEmotionDims(boolean emodims)

Method to consider emotion dimensions.

Parameters:

emodims - The emotion dimensions flag.

setNegation

public void setNegation(boolean neg)

Method to consider negations.

Parameters:

neg - The negation flag.

setPrior

public void setPrior(java.lang.String prior)

Method to set the regression prior.

Parameters:

prior - The regression prior.

save

public void save(java.lang.String path)

Description copied from class: Classifier

Generic method to save the fully fledged classifier into a given file path. It is recommended to use a plain text file (such as XML) to save the classifier's configuration since it's readable directly.

Specified by:

save in class Classifier

Parameters:

path - The file path to save the classifier.

load

public void load(java.lang.String path)

Description copied from class: Classifier

Generic function to load a previously saved classifier. This function should be consistent with the design followed in the saving procedure.

Specified by:

load in class Classifier

Parameters:

path - The path of the file which contains the previously saved classifier.

resetExamples

public void resetExamples()

Description copied from class: Classifier

Method to reset the classifier and flush the training examples. This method only makes sense if the classifier in question is trainable and already has some training examples.

Overrides:

resetExamples in class Classifier

main

public static void main(java.lang.String[] args)
                 throws java.lang.Exception

Main method to test the LR classifier.

Throws:

java.lang.Exception