Bayesian Estimation for Pair wise Markov Chains in Signal Processing

In this paper we study Pair wise Markov Chain (PMC), which generalizes the classical Hidden Markov Chain (HMC) model. This generalization allows one to model more complex situations in signal processing etc. In a PMC the hidden process is not necessarily a Markov process. However, PMC allows one to use the classical Bayesian estimation techniques of Highest Posterior Density (HPD) Pair wise Markov Chains allows one to restore hidden random processes, similar to HMC, with major applications in signal processing. The above is illustrated with numerical examples.