Algorithmic Issues in Hidden Markov Models

David Fernández-Baca (1)

e-mails: fernande@cs.iastate.edu

(1) Department of Computer Science, Iowa State University, Ames, Iowa 50011 Estados Unidos

Abstract

A hidden Markov model (HMM) is a stochastic system that can, at any given time, be in one of a finite number of states, each of which emits a symbol with a certain probability. Furthermore, transitions between states occur according to certain probabilities. An observer of a HMM can see the sequence it emits, but not the sequence of states that produced the symbols. A basic problem in HMMs is to infer the most likely sequence of states that resulted in a given observed sequence. HMMs are used in applications ranging from speech recognition to gene identification. For example, in speech recognition the observed symbols are a series of phonemes and the problem is to infer the sequence of words that produced it. In this tutorial, we give an overview of HMMs and discuss some of their applications, with special emphasis on their use in bioinformatics. We then discuss some of the algorithmic issues that arise in conjunction with HMMs. In particular, we consider methods for studying the sensitivity of HMMs to the choice of transition probabilities and for estimating the best parameters for a model. We illustrate these approaches through an application in computational biology: estimating the evolutionary distance between two DNA sequences.

Keywords:Hidden Markov models, Statistical models, Bioinformatics, Computational biology, Algorithms, Evolutionary trees, Sensitivity analysis, Optimization

Biografía/Biography

David Fernández-Baca is a Professor of Computer Science at Iowa State University, where he has been a faculty member since 1986. He obtained the undergraduate degree in Computer Engineering (Ingeniería en Computación) in 1980 from the Universidad Nacional Autónoma de México, the MS in Computer Engineering and the PhD in Computer Science from the University of California, Davis in 1983 and 1986, respectively. His research interests are in computational biology (primarily in evolutionary tree construction) and combinatorial optimization (primarily in sensitivity analysis of optimization problems).


BibTex

@INPROCEEDINGS{fernandez-baca04:1008,
                  AUTHOR       = {David Fernández-Baca},
                  TITLE        = {Algorithmic Issues in Hidden Markov Models},
                  BOOKTITLE    = {30ma Conferencia Latinoamericana de Informática (CLEI2004)},
                  YEAR         = {2004},
                  editor       = {Mauricio Solar and David Fernández-Baca and Ernesto Cuadros-Vargas},
                  pages        = {12--12},
                  address      = {},
                  month        = Sep,
                  organization = {Sociedad Peruana de Computación},
                  note         = {ISBN 9972-9876-2-0},
}

PDF de CLEI2004 (incluye todos los artículos)
Página principal CLEI 2004
Generado por Sociedad Peruana de Computación