GPD-based state modification by weighted linear loss function

Taehee Kwon, Hanseok Ko

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

This paper proposes an effective loss function to assign the HMM state weight based on the MCE method. This is to remedy the performance limitation inherent in the traditional maximum likelihood method, which adjusts parameters to maximize the likelihood of training HMM. If minimum classification error method is used to minimize the error rate for training data, then the local optimum point can be achieved and the recognition performance can be achieved. However, if the amount of data used in the MCE training is too small, there can be a risk of overfilling to the training data. In this paper, we propose state modification by weighted linear loss function to overcome overfilling to training data. Representative experiments confirm this postulation and show the improvement in error rate when applied.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
EditorsPeter M. A. Sloot, David Abramson, Alexander V. Bogdanov, Yuriy E. Gorbachev, Jack J. Dongarra, Albert Y. Zomaya
PublisherSpringer Verlag
Pages1100-1108
Number of pages9
ISBN (Print)3540401970, 9783540401971
DOIs
Publication statusPublished - 2003

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2660
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'GPD-based state modification by weighted linear loss function'. Together they form a unique fingerprint.

Cite this