Baum-Welch Learning in Discrete Hidden Markov Models with Linear Factorial Constraints

Autor:	Jens R. Otterpohl
Rok vydání:	2002
Předmět:	Factorial symbols.namesake Transformation (function) symbols Forward algorithm Hidden semi-Markov model Cartesian product Markov model Hidden Markov model Baum–Welch algorithm Algorithm Mathematics
Zdroj:	Artificial Neural Networks — ICANN 2002 ISBN: 9783540440741 ICANN
DOI:	10.1007/3-540-46084-5_191
Popis:	Here, I introduce a transformation-based method for extending the Baum-Welch algorithm to the training of discrete Hidden Markov Models subject to constraints on the parameters. A class of certain linear factorial constraints is described and shown to lead to exact reestimation formulas. Applying these constraints to the hidden state transitions allows to estimate processes that are cartesian products of multiple subprocesses on differing timescales. The applicability of the method has been demonstrated previously using constraints on both hidden and observation processes. The potential benefit of the approach is discussed in qualitative comparison to factorial Hidden Markov Model architectures.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::cb7d5ecb9f688c5f098eec01190b670b https://doi.org/10.1007/3-540-46084-5_191 Zobrazit plný text záznamu