it.uniroma2.sag.kelp.learningalgorithm.classification.hmm

Class SequenceClassificationKernelBasedLearningAlgorithm

java.lang.Object

it.uniroma2.sag.kelp.learningalgorithm.classification.hmm.SequenceClassificationLearningAlgorithm

it.uniroma2.sag.kelp.learningalgorithm.classification.hmm.SequenceClassificationKernelBasedLearningAlgorithm

All Implemented Interfaces:

KernelMethod, LearningAlgorithm, MetaLearningAlgorithm

public class SequenceClassificationKernelBasedLearningAlgorithm
extends SequenceClassificationLearningAlgorithm
implements KernelMethod

/** This class implements a sequential labeling paradigm.
Given sequences of items (each implemented as an Example and associated to one Label) this class allow to apply a generic LearningAlgorithm to use the "history" of each item in the sequence in order to improve the classification quality. In other words, the classification of each example does not depend only its representation, but it also depend on its "history", in terms of the classed assigned to the preceding examples.
This class should be used when a kernel-based learning algorithm is used, thus directly operating in the implicit space underlying a kernel function.
This algorithms was inspired by the work of:
Y. Altun, I. Tsochantaridis, and T. Hofmann. Hidden Markov support vector machines. In Proceedings of the Twentieth International Conference on Machine Learning, 2003.

Author:

Danilo Croce

Constructor Summary

Constructors

Constructor and Description
SequenceClassificationKernelBasedLearningAlgorithm()
SequenceClassificationKernelBasedLearningAlgorithm(BinaryLearningAlgorithm baseLearningAlgorithm, int transitionsOrder, float transitionWeight)

Method Summary

Methods

Modifier and Type	Method and Description
LearningAlgorithm	duplicate() Creates a new instance of the LearningAlgorithm initialized with the same parameters of the learningAlgorithm to be duplicated.
LearningAlgorithm	getBaseAlgorithm() Returns the base algorithm this meta algorithm is based on
Kernel	getKernel() Returns the kernel exploited by this learner
void	setBaseAlgorithm(LearningAlgorithm baseAlgorithm)
void	setKernel(Kernel kernel) Sets the kernel this
void	setKernelCache(KernelCache cache)

Methods inherited from class it.uniroma2.sag.kelp.learningalgorithm.classification.hmm.SequenceClassificationLearningAlgorithm

getBaseLearningAlgorithm, getBeamSize, getLabels, getMaxEmissionCandidates, getPredictionFunction, getSequenceExampleGenerator, getTransitionsOrder, learn, reset, setBaseLearningAlgorithm, setBeamSize, setLabels, setMaxEmissionCandidates, setPredictionFunction, setSequenceExampleGenerator

Methods inherited from class java.lang.Object

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

Constructor Detail

SequenceClassificationKernelBasedLearningAlgorithm

public SequenceClassificationKernelBasedLearningAlgorithm()

SequenceClassificationKernelBasedLearningAlgorithm

public SequenceClassificationKernelBasedLearningAlgorithm(BinaryLearningAlgorithm baseLearningAlgorithm,
                                                  int transitionsOrder,
                                                  float transitionWeight)
                                                   throws Exception

Parameters:

baseLearningAlgorithm - the learning algorithm devoted to the acquisition of a model after that each example has been enriched with its "history"

transitionsOrder - given a targeted item in the sequence, this variable determines the number of previous example considered in the learning/labeling process.

transitionWeight - the importance of the transition-based features during the learning process. Higher valuers will assign more importance to the transitions.

Throws:

Exception - The input baseLearningAlgorithm is not a kernel-based method

Method Detail

duplicate

public LearningAlgorithm duplicate()

Description copied from interface: LearningAlgorithm

Creates a new instance of the LearningAlgorithm initialized with the same parameters of the learningAlgorithm to be duplicated. A shallow copy of the learning parameters is performed, while the state parameters are initialized to their default value. The labels are not initialized neither copied

Specified by:

duplicate in interface LearningAlgorithm

getBaseAlgorithm

public LearningAlgorithm getBaseAlgorithm()

Description copied from interface: MetaLearningAlgorithm

Returns the base algorithm this meta algorithm is based on

Specified by:

getBaseAlgorithm in interface MetaLearningAlgorithm

Returns:

the base algorithm

getKernel

public Kernel getKernel()

Description copied from interface: KernelMethod

Returns the kernel exploited by this learner

Specified by:

getKernel in interface KernelMethod

Returns:

the kernel

setBaseAlgorithm

public void setBaseAlgorithm(LearningAlgorithm baseAlgorithm)

Specified by:

setBaseAlgorithm in interface MetaLearningAlgorithm

Parameters:

baseAlgorithm - the baseAlgorithm to set

setKernel

public void setKernel(Kernel kernel)

Description copied from interface: KernelMethod

Sets the kernel this

Specified by:

setKernel in interface KernelMethod

Parameters:

kernel - the kernel to set

setKernelCache

public void setKernelCache(KernelCache cache)