Please use this identifier to cite or link to this item:
Title: A Nonstationary Hidden Markov Model with Approximately Infinitely-Long Time-Dependencies
Authors: Chatzis, Sotirios P. 
Kosmopoulos, Dimitrios 
Papadourakis, George M. 
Keywords: Pattern Recognition;Computer Graphics;Image processing and computer vision;User interfaces and human computer interaction;Information systems applications (incl. Internet);Computational biology/bioinformatics
Category: Computer and Information Sciences
Field: Engineering and Technology
Issue Date: Dec-2014
Publisher: Springer International Publishing
Source: International Symposium on Visual Computing ISVC 2014: Advances in Visual Computing, pp. 51-62
Conference: International Symposium on Visual Computing (ISVC) 
Abstract: Hidden Markov models (HMMs) are a popular approach for modeling sequential data, typically based on the assumption of a first-order Markov chain. In other words, only one-step back dependencies are modeled which is a rather unrealistic assumption in most applications. In this paper, we propose a method for postulating HMMs with approximately infinitely-long time-dependencies. Our approach considers the whole history of model states in the postulated dependencies, by making use of a recently proposed nonparametric Bayesian method for modeling label sequences with infinitely-long time dependencies, namely the sequence memoizer. We manage to derive training and inference algorithms for our model with computational costs identical to simple first-order HMMs, despite its entailed infinitely-long time-dependencies, by employing a mean-field-like approximation. The efficacy of our proposed model is experimentally demonstrated.
ISBN: 978-3-319-14363-7
978-3-319-14364-4 (online)
DOI: 10.1007/978-3-319-14364-4_6
Rights: © Springer
Type: Conference Papers
Appears in Collections:Δημοσιεύσεις σε συνέδρια/Conference papers

Show full item record

Page view(s)

Last Week
Last month
checked on Oct 18, 2019

Google ScholarTM



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