A Hidden Markov Model (HMM) is a common statistical model which is widely used for analysis of biological sequence data and other sequential phenomena. In the present paper we extend HMMs with constraints and show how the familiar Viterbi algorithm can be generalized, based on constraint solving methods. HMMs with constraints have advantages over traditional ones in terms of more compact expressions as well as opportunities for pruning during Viterbi computations. We exemplify this by an enhancement of a simple prokaryote gene finder given by an HMM.
Proceedings of Wcb09: Workshop on Constraint Based Methods for Bioinformatics, 2009
Hidden Markov Model; Constraint Programming; Constrained Hidden Markov Model
Main Research Area:
Workshop on Constraint Based Methods for Bioinformatics, 2009