Fast estimation of Hidden Markov Models via alpha-EM algorithm

Yasuo Matsuyama*, Ryunosuke Hayashi, Ryota Yokote

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contribution


    Fast estimation algorithms of Hidden Markov Models (HMMs), or alpha-HMMs, are presented. Such novel algorithms inherit speedup properties of the alpha-EM algorithm. Since the alpha-EM algorithm includes the traditional log-EM algorithm as its special case, the alpha-HMM also includes the traditional log-HMM as its special case. This generalization appears as the utilization of the past information which is the main device of the speedup. Since the memorization of the past information requires only little increase of computational load and memory, the iteration speedup directly appears as that of CPU time. Experimental results are given.

    Original languageEnglish
    Title of host publicationIEEE Workshop on Statistical Signal Processing Proceedings
    Number of pages4
    Publication statusPublished - 2011
    Event2011 IEEE Statistical Signal Processing Workshop, SSP 2011 - Nice
    Duration: 2011 Jun 282011 Jun 30


    Other2011 IEEE Statistical Signal Processing Workshop, SSP 2011


    • alpha-EM algorithm
    • alpha-HMM
    • past information
    • speedup

    ASJC Scopus subject areas

    • Electrical and Electronic Engineering
    • Applied Mathematics
    • Signal Processing
    • Computer Science Applications


    Dive into the research topics of 'Fast estimation of Hidden Markov Models via alpha-EM algorithm'. Together they form a unique fingerprint.

    Cite this