MAP segmentation in Bayesian hidden Markov models: a case study
This work addresses segmentation in HMMs for bioinformatics applications, but it is incremental as it applies existing Bayesian methods to a specific dataset without major methodological breakthroughs.
The paper tackles the problem of estimating the maximum posterior probability state sequence in Bayesian hidden Markov models with Dirichlet priors, using protein alignment data to set hyperparameters and comparing iterative algorithms, finding that the Bayesian setup performs competitively against the frequentist approach in this case study.
We consider the problem of estimating the maximum posterior probability (MAP) state sequence for a finite state and finite emission alphabet hidden Markov model (HMM) in the Bayesian setup, where both emission and transition matrices have Dirichlet priors. We study a training set consisting of thousands of protein alignment pairs. The training data is used to set the prior hyperparameters for Bayesian MAP segmentation. Since the Viterbi algorithm is not applicable any more, there is no simple procedure to find the MAP path, and several iterative algorithms are considered and compared. The main goal of the paper is to test the Bayesian setup against the frequentist one, where the parameters of HMM are estimated using the training data.