prob_evidence.m

来自「Bayes网络工具箱」· M 代码 · 共 17 行

function loglik = prob_evidence(engine, evidence)
% PROB_EVIDENCE Compute the likelihood of the observation sequence (hmm)
% loglik = prob_evidence(engine, evidence)
%
% evidence{i,t} = [] if if X(i,t) is hidden, and otherwise contains its observed value (scalar or column vector)

bnet = bnet_from_engine(engine);
n = length(bnet.intra);
onodes = engine.onodes;
T = size(evidence, 2);
assert(~any(isemptycell(evidence(onodes,:))));

obslik = mk_hmm_obs_lik_mat(bnet, onodes, evidence);

filter = 1;
[alpha, beta, gamma, loglik] = forwards_backwards_oneslice(engine.prior, engine.transmat, obslik, filter);