Please use this identifier to cite or link to this item: http://ktisis.cut.ac.cy/handle/10488/7259
Title: A variational Bayesian methodology for hidden Markov models utilizing Student's-t mixtures
Authors: Chatzis, Sotirios P. 
Kosmopoulos, Dimitrios 
Chatzis, Sotirios P. 
Kosmopoulos, Dimitrios 
Keywords: Pattern recognition
Markov processes
Human-computer interaction
Loudspeakers
Robotics
Robots
Issue Date: 2011
Publisher: Elsevier
Source: Pattern recognition, 2011, Volume 44, Issue 2, Pages 295–306
Abstract: The Student's-t hidden Markov model (SHMM) has been recently proposed as a robust to outliers form of conventional continuous density hidden Markov models, trained by means of the expectationmaximization algorithm. In this paper, we derive a tractable variational Bayesian inference algorithm for this model. Our innovative approach provides an efficient and more robust alternative to EM-based methods, tackling their singularity and overfitting proneness, while allowing for the automatic determination of the optimal model size without cross-validation. We highlight the superiority of the proposed model over the competition using synthetic and real data. We also demonstrate the merits of our methodology in applications from diverse research fields, such as human computer interaction, robotics and semantic audio analysis
URI: http://ktisis.cut.ac.cy/handle/10488/7259
ISSN: 00313203
DOI: 10.1016/j.patcog.2010.09.001
Rights: © Elsevier Ltd. All rights reserved
Appears in Collections:Άρθρα/Articles

Show full item record

SCOPUSTM   
Citations 20

17
checked on Mar 27, 2017

WEB OF SCIENCETM
Citations 20

13
checked on Mar 27, 2017

Page view(s)

11
checked on Apr 26, 2017

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.