Package de.jstacs.sequenceScores.statisticalModels.trainable.hmm.transitions

The package provides all interfaces and classes for transitions used in hidden Markov models.

See:
Description

Interface Summary

DifferentiableTransition	This class declares methods that allow for optimizing the parameters numerically using the Optimizer.
SamplingTransition	This interface declares all method used during a sampling.
TrainableAndDifferentiableTransition	This interface unifies the interfaces TrainableTransition and DifferentiableTransition.
TrainableTransition	This class declares methods that allow for estimating the parameters from a sufficient statistic, as for instance done in the (modified) Baum-Welch algorithm, viterbi training, or Gibbs sampling.
Transition	This interface declares the methods of the transition used in a hidden Markov model.
TransitionWithSufficientStatistic	This interface defines method for reseting and filling an internal sufficient statistic.

Class Summary

BasicHigherOrderTransition	This class implements the basic transition that allows to be trained using the viterbi or the Baum-Welch algorithm.
BasicHigherOrderTransition.AbstractTransitionElement	This class declares the probability distribution for a given context, i.e. it contains all possible transition and the corresponding probabilities for a given set offset previously visited states.
HigherOrderTransition	This class can be used in any AbstractHMM allowing to use gradient based or sampling training algorithm.

Package de.jstacs.sequenceScores.statisticalModels.trainable.hmm.transitions Description

The package provides all interfaces and classes for transitions used in hidden Markov models. Each transition is based on a set of BasicHigherOrderTransition.AbstractTransitionElements.

Since:

Jstacs 1.5

See Also:

Transition, BasicHigherOrderTransition.AbstractTransitionElement, de.jstacs.sequenceScores.statisticalModels.trainable.hmm.transitions.elements