A high-order hidden Markov model is proposed for better approximating the real process.

What is it about?

An expectation-maximization based algorithm was presented for the parameter estimation of the proposed model. In this study, we also apply the proposed model to noisy speech recognition.

Featured Image

Why is it important?

The hidden Markov models (HMMs) have been widely applied to systems that deal with sequential data. However, the conditional independence of the state outputs will limit the output of a hidden Markov model to be a piecewise constant random sequence, which is not a good approximation for many real processes. In this study, the proposed model is capable of better approximating the real process due to get rid of the limit that the state outputs are constrained to be conditional independence.