Filtering of hidden weak markov chain -discrete range observations
EM algorithm, Filtering, Hidden weak Markov chain, Parameter reestimation, Smoothing
International Series in Operations Research and Management Science
In this paper we consider a hidden discrete time finite state process X whose behavior at the present time t depends on its behavior at the previous k time steps, which is a generalization of the usual hidden finite state Markov chain, in which k equals to one. We consider the case when the range space of our observations is finite. We present filtering equations for certain functionals of the chain and perform related error analysis.
Original Publication Date
Luo, Shangzhen and Tsoi, Allanus H., "Filtering of hidden weak markov chain -discrete range observations" (2007). Faculty Publications. 2676.