Sequence Clustering with Discrete-Output HMMs
Cluster Assignment
HMM BIC
Count HMM Parameters
Heatmap Emission Probabilities
DBHC Algorithm
Get HMM Log Likelihood
Partition BIC
Seed Selection Procedure
Sequence-to-HMM Likelihood
Size Search Algorithm
Smooth HMM Parameters
Smooth Probabilities
Heatmap Transition Probabilities
Provides an implementation of a mixture of hidden Markov models (HMMs) for discrete sequence data in the Discrete Bayesian HMM Clustering (DBHC) algorithm. The DBHC algorithm is an HMM Clustering algorithm that finds a mixture of discrete-output HMMs while using heuristics based on Bayesian Information Criterion (BIC) to search for the optimal number of HMM states and the optimal number of clusters.