Computation of Entropy of Most Likelihood State Sequence Obtained from Non Homogeneous Fuzzy Hidden Markov Chain

Abstract

The entropy of a possibilistic variable provides a measure of its uncertainty. An algorithm is proposed for computing the entropy of the most likelihood state sequence obtained from the Viterbi algorithm for Non Homogeneous Fuzzy Hidden Markov Chain (NHFHMC) which is a bivariate discrete process, where is a non homogeneous fuzzy Markov chain on possibility space and is the sequence of observations such that the conditional possibility distribution of only depends on [8]. The Viterbi algorithm for NHFHMC is the algorithm for tracking the most likelihood hidden states of a process from a sequence of observations. An important problem while tracking a process is estimating the uncertainty present in the solution. To overcome this kind of uncertainty we have computed the entropy associated with that most likelihood state sequence and this entropy measure is given in triangular fuzzy number.