A new algorithm for inference in HMM's with lower span complexity

Diogo Pereira, Cláudia Nunes, Rui Rodrigues

Research output: Contribution to journalArticlepeer-review

3 Citations (Scopus)
9 Downloads (Pure)

Abstract

The maximum likelihood problem for Hidden Markov Models is usually numerically solved by the Baum-Welch algorithm, which uses the Expectation-Maximization algorithm to find the estimates of the parameters. This algorithm has a recursion depth equal to the data sample size and cannot be computed in parallel, which limits the use of modern GPUs to speed up computation time. A new algorithm is proposed that provides the same estimates as the Baum-Welch algorithm, requiring about the same number of iterations, but is designed in such a way that it can be parallelized. As a consequence, it leads to a significant reduction in the computation time. This reduction is illustrated by means of numerical examples, where we consider simulated data as well as real datasets.

Original languageEnglish
Article number107955
JournalComputational Statistics and Data Analysis
Volume195
DOIs
Publication statusPublished - Jul 2024

Keywords

  • Baum-Welch algorithm
  • Expectation-Maximization algorithm
  • Hidden Markov Models
  • Parallel computation

Cite this