ph_em.m

来自「隐马尔可夫工具箱」· M 代码 · 共 13 行

function [TRANS, rate, logl, beta, dens] = ph_em(count, TRANS, rate, Nit)

%ph_em	  Estimates the parameters of a Poisson HMM using the EM algorithm.
%         Use: [TRANS,rate,logl,postprob,dens] =
%         ph_em(count,TRANS_0,rate_0,Nit). where TRANS and rate are
%         the estimated model parameters, logl contains the
%         log-likehood values for the successive iterations, postprob
%         the marginal posterior probabilities at the last iteration
%         and dens the negative binomial probabilities computed also
%         at the last iteration.

% H2M/cnt Toolbox, Version 2.0
% Olivier Capp