Hidden Markov Models--Model selection and checking
建议楼主使用package"HiddenMarkov"
function 较齐全
function logLik() 可算出 Log Likelihood of Hidden Markov Model
然后依公式就可算出 AIC or BIC
1. Akaike's information criteria: AIC = -2log L + 2p.
2. Bayesian information criteria: BIC = -2log L + plogN,
where L is the likelihood of the fitted model,
p is the number of parameters,
and N is the number of data points.