HmmTrainByEM (SuanShu 1.3.1 API Documentation)

Overview

Package

Class

Tree

Index

Help

SuanShu, a Java numerical and statistical library

PREV CLASS NEXT CLASS

FRAMES NO FRAMES

SUMMARY: NESTED | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

com.numericalmethod.suanshu.stats.hmm
Class HmmTrainByEM

java.lang.Object
  com.numericalmethod.suanshu.stats.hmm.HmmTrainByEM

This class trains a HMM model by observations using the train algorithm.

Constructor Summary
`HmmTrainByEM()`

Method Summary
`HiddenMarkovModel`	`train(HiddenMarkovModel model0, int[] observations)` Construct a trained (discrete) hidden Markov model using the train algorithm.

Methods inherited from class java.lang.Object
`clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait`

Constructor Detail

public HmmTrainByEM()

Method Detail

public HiddenMarkovModel train(HiddenMarkovModel model0,
                               int[] observations)

Construct a trained (discrete) hidden Markov model using the train algorithm.

Parameters:: model0 - the original hidden Markov model; observations - an integer array of observation symbols (length = T); each symbol should be a positive integer less than or equal to M (the number of district observation symbols per state)
Returns:: a trained (discrete) hidden Markov model