High-order hidden Markov models - Estimation and implementation

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

While the Hidden Markov Model (HMM) has been used for a wide range of applications, an efficient procedure for estimating the model parameters and finding the optimal state sequence has not been generally formulated for orders higher than first, i.e., for models that assume higher order of either the state sequence memory, or the dependency of the observations on the states. We propose a simple method that transforms any high order HMM (including models in which the state sequence and observation dependency are of different orders) into an equivalent first order one, and thus makes the first order HMM formulation applicable to models of any order.

Original languageEnglish
Title of host publication2009 IEEE/SP 15th Workshop on Statistical Signal Processing, SSP '09
Pages249-252
Number of pages4
DOIs
StatePublished - 2009
Event2009 IEEE/SP 15th Workshop on Statistical Signal Processing, SSP '09 - Cardiff, United Kingdom
Duration: 31 Aug 20093 Sep 2009

Publication series

NameIEEE Workshop on Statistical Signal Processing Proceedings

Conference

Conference2009 IEEE/SP 15th Workshop on Statistical Signal Processing, SSP '09
Country/TerritoryUnited Kingdom
CityCardiff
Period31/08/093/09/09

Fingerprint

Dive into the research topics of 'High-order hidden Markov models - Estimation and implementation'. Together they form a unique fingerprint.

Cite this