de.jstacs.models.hmm.training
Class MaxHMMTrainingParameterSet

java.lang.Object
  de.jstacs.parameters.ParameterSet
      de.jstacs.models.hmm.HMMTrainingParameterSet
          de.jstacs.models.hmm.training.MaxHMMTrainingParameterSet

All Implemented Interfaces:

GalaxyConvertible, RangeIterator, Storable, Cloneable

Direct Known Subclasses:

BaumWelchParameterSet, NumericalHMMTrainingParameterSet, ViterbiParameterSet

public abstract class MaxHMMTrainingParameterSet
extends HMMTrainingParameterSet

This class is the super class for any HMMTrainingParameterSet that is used for a maximizing training algorithm of a hidden Markov model.

Author:

Jens Keilwagen

Nested Class Summary

Nested classes/interfaces inherited from class de.jstacs.parameters.ParameterSet

ParameterSet.ParameterList

Field Summary

Fields inherited from class de.jstacs.parameters.ParameterSet

errorMessage, parameters, parent, ranged

Constructor Summary

protected	MaxHMMTrainingParameterSet() This is the empty constructor that can be used to fill the parameters after creation.
protected	MaxHMMTrainingParameterSet(int starts, AbstractTerminationCondition tc) This constructor can be used to create an instance with specified parameters.
protected	MaxHMMTrainingParameterSet(StringBuffer xml) The standard constructor for the interface Storable.

Method Summary

AbstractTerminationCondition	getTerminantionCondition() This method returns the AbstractTerminationCondition for stopping the training, e.g., if the difference of the scores between two iterations is smaller than a given threshold the training is stopped.
protected void	loadParameters() Loads the parameters for this ParameterSet.

Methods inherited from class de.jstacs.models.hmm.HMMTrainingParameterSet

getNumberOfStarts

Methods inherited from class de.jstacs.parameters.ParameterSet

clone, fromGalaxy, fromXML, getErrorMessage, getId, getIndex, getNumberOfParameters, getNumberOfValues, getParameterAt, getParent, hasDefaultOrIsSet, initParameterList, initParameterList, isAtomic, isRanged, makeRanged, next, parametersLoaded, propagateId, recieveId, replaceParametersWithRangedInstance, reset, resetToFirst, setParent, simplify, toGalaxy, toXML, valuesToString

Methods inherited from class java.lang.Object

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

Constructor Detail

MaxHMMTrainingParameterSet

protected MaxHMMTrainingParameterSet()

This is the empty constructor that can be used to fill the parameters after creation.

MaxHMMTrainingParameterSet

protected MaxHMMTrainingParameterSet(int starts,
                                     AbstractTerminationCondition tc)
                              throws Exception

This constructor can be used to create an instance with specified parameters.

Parameters:

starts - the number of different starts

tc - the termination condition for stopping the algorithm

Throws:

Exception - if this MaxHMMTrainingParameterSet could not be created

MaxHMMTrainingParameterSet

protected MaxHMMTrainingParameterSet(StringBuffer xml)
                              throws NonParsableException

The standard constructor for the interface Storable. Constructs a MaxHMMTrainingParameterSet out of an XML representation.

Parameters:

xml - the XML representation as StringBuffer

Throws:

NonParsableException - if the MaxHMMTrainingParameterSet could not be reconstructed out of the StringBuffer xml

Method Detail

loadParameters

protected void loadParameters()
                       throws Exception

Description copied from class: ParameterSet

Loads the parameters for this ParameterSet. This is in most cases done by calling ParameterSet.initParameterList() or ParameterSet.initParameterList(int) to initialize ParameterSet.parameters and afterwards filling ParameterSet.parameters with instances of subclasses of Parameter.

Overrides:

loadParameters in class HMMTrainingParameterSet

Throws:

Exception - if the parameters could not be loaded

See Also:

Parameter, ParameterSet.parameters, ParameterSet.initParameterList(), ParameterSet.initParameterList(int)

getTerminantionCondition

public AbstractTerminationCondition getTerminantionCondition()
                                                      throws ParameterSetParser.NotInstantiableException

This method returns the AbstractTerminationCondition for stopping the training, e.g., if the difference of the scores between two iterations is smaller than a given threshold the training is stopped.

Returns:

the AbstractTerminationCondition for stopping the training

Throws:

ParameterSetParser.NotInstantiableException - if the AbstractTerminationCondition could not be created from its ParameterSet