Churbanov Linear Memory Backwards Pass Baum Welch HMM Optimisation Intended for use with BGHMM.jl in distributed training of large model "zoos" of background genomic hidden markov models. Allows observations arrays with multiple observations sequences (intended for categorical dists only- end of sequence is indexed by a trailing zero 0).
mmattocks / clhmm.jl Goto Github PK
View Code? Open in Web Editor NEWChurbanov Linear Memory Baum-Welch HMM Optimisation for Background Genomic HMMs
License: MIT License