hmm-class {tileHMM}R Documentation

Class "hmm"

Description

Virtual base class for HMMs.

Objects from the Class

Do not create objects of this class directly. Instead use objects of derived classes like "contHMM".

Slots

transition.matrix:
Object of class "matrix", storing the transition probabilities of the Markov chain.
emission:
Object of class "list" containing objects of class "dist" to represent emission distributions for each state.
init:
Object of class "numeric". The initial state distribution of the Markov chain.

Methods

baumWelch
signature(hmm = "hmm", obs = "list"): Baum-Welch algorithm to estimate model parameters.
viterbiTraining
signature(hmm = "hmm", obs = "list"): Viterbi training to estimate model parameters.
viterbi
signature(hmm = "hmm"): Viterbi algorithm to calculate most likely state sequence.
backward
signature(hmm = "hmm"): Computing of backward variables.
forward
signature(hmm = "hmm"): Computing of forward variables.
states
signature(hmm = "hmm"): Returns state names.

Author(s)

Peter Humburg

See Also

contHMM, baumWelch, viterbiTraining, viterbi, forward, backward, states

Examples

showClass("hmm")

[Package tileHMM version 1.0-2 Index]