Current browse context:
math.CO
Change to browse by:
References & Citations
Mathematics > Combinatorics
Title: Identifying combinatorially symmetric Hidden Markov Models
(Submitted on 9 Sep 2017)
Abstract: We provide a sufficient criterion for the unique parameter identification of combinatorially symmetric Hidden Markov Models based on the structure of their transition matrix. If the observed states of the chain form a zero forcing set of the graph of the Markov model then it is uniquely identifiable and an explicit reconstruction method is given.
Link back to: arXiv, form interface, contact.