Package dev.nm.stat.hmm.discrete

Class DiscreteHMM

java.lang.Object

dev.nm.stat.markovchain.SimpleMC

dev.nm.stat.hmm.HMMRNG

dev.nm.stat.hmm.HiddenMarkovModel

dev.nm.stat.hmm.discrete.DiscreteHMM

All Implemented Interfaces:

RandomNumberGenerator, Seedable

Direct Known Subclasses:

BaumWelch

public class DiscreteHMM
extends HiddenMarkovModel

This is the discrete hidden Markov model as defined by Rabiner.

Constructor Summary

Constructors

Constructor	Description
DiscreteHMM(Vector PI, Matrix A, Matrix B)	Constructs a discrete hidden Markov model.
DiscreteHMM(DiscreteHMM model)	Copy constructor.

Method Summary

Modifier and Type	Method	Description
ImmutableMatrix	B()	Gets the conditional probabilities of the observation symbols: rows correspond to state; columns corresponds symbols.
double	density(int state, double observation)	Gets the (conditional) probability mass of making an observation in a particular state.
int	nSymbols()	Gets the number of observation symbols per state.

Methods inherited from class dev.nm.stat.hmm.HiddenMarkovModel

logProbability, logProbability, logProbability

Methods inherited from class dev.nm.stat.hmm.HMMRNG

next, nextDouble, seed

Methods inherited from class dev.nm.stat.markovchain.SimpleMC

A, bin, getStationaryProbabilities, nextState, nStates, PI

Methods inherited from class java.lang.Object

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

Constructor Details

DiscreteHMM

public DiscreteHMM(Vector PI,
 Matrix A,
 Matrix B)

Constructs a discrete hidden Markov model.

Parameters:

PI - the initial state probabilities

A - the state transition probabilities

B - the conditional probabilities of the observation symbols: rows correspond to state; columns corresponds symbols

DiscreteHMM

public DiscreteHMM(DiscreteHMM model)

Copy constructor.

Parameters:

model - a HiddenMarkovModel

Method Details

B

public ImmutableMatrix B()

Gets the conditional probabilities of the observation symbols: rows correspond to state; columns corresponds symbols.

Returns:

the observation symbol probabilities

nSymbols

public int nSymbols()

Gets the number of observation symbols per state.

Returns:

the number of observation symbols per state

density

public double density(int state,
 double observation)

Gets the (conditional) probability mass of making an observation in a particular state.

Specified by:

density in class HiddenMarkovModel

Parameters:

state - the hidden state label, counting from 1

observation - the observation value

Returns:

the probability density/mass