A note on estimation of states in HMM with unknown parameters
Yasunari Maeda, Fumito Masui, Masakiyo Suzuki · 2011
Summary Estimation of states in HMM(Hidden Markov Model) with unknown parameters is very important topic in many fields. In this paper we propose an optimal estimation method which minimizes an error rate with reference to the Bayes criterion. We also propose approximate method in order to reduce computational complexity.