PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Adjusted Viterbi training
Alexey Koloydenko and Jüri Lember
(2003) Eurandom, Eindhoven, The Netherlands.

Abstract

We propose modifications of the Viterbi Training (VT) algorithm to estimate emission parameters in Hidden Markov Models (HMM) which are widely used in speech recognition, natural language modeling, image analysis, and bioinformatics. Our goal is to alleviate the inconsistency of VT while controlling the amount of extra computations. Specifically, we modify VT to enable it asymptotically to fix the true values of the parameters as does the EM algorithm. Our approach relies on infinite Viterbi alignment and an associated with it limiting probability distribution. We focus on mixture models, an important special case of HMM, wherein the limiting distribution can be computed exactly and be used in the adjusted VT algorithm. A simulation experiment shows that our central algorithm (VA1) can dramatically improve accuracy without much cost in computation time. We also propose VA2, a more mathematically advanced correction to VT, verify its fast convergence and high accuracy, and intend to elaborate on its computationally feasible implementations in future work.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Other
Additional Information:Submitted to Computational Statistics and Data Analysis
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Learning/Statistics & Optimisation
Speech
ID Code:247
Deposited By:Alexey Koloydenko
Deposited On:23 November 2004