Probabilistic Kernel Principal Component Analysis Through Time

Mauricio Alvarez, Ricardo Henao

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

This paper introduces a temporal version of Probabilistic Kernel Principal Component Analysis by using a hidden Markov model in order to obtain optimized representations of observed data through time. Recently introduced, Probabilistic Kernel Principal Component Analysis overcomes the two main disadvantages of standard Principal Component Analysis, namely, absence of probability density model and lack of high-order statistical information due to its linear structure. We extend this probabilistic approach of KPCA to mixture models in time, to enhance the capabilities of transformation and reduction of time series vectors. Results over voice disorder databases show improvements in classification accuracies even with highly reduced representations.
Original languageEnglish
Title of host publicationLecture Notes in Computer Science : Neural Information Processing
Volume4232
Place of PublicationBerlin
PublisherSpringer
Publication date2006
Pages747-754
ISBN (Print)978-3-540-46479-2
DOIs
Publication statusPublished - 2006
Externally publishedYes
Event13th International Conference on Neural Information Processes - Hong Kong, China
Duration: 3 Oct 20066 Oct 2006
Conference number: 13
http://www.informatik.uni-trier.de/~ley/db/conf/iconip/iconip2006-3.htmlc

Conference

Conference13th International Conference on Neural Information Processes
Number13
CountryChina
CityHong Kong
Period03/10/200606/10/2006
Internet address
SeriesLecture Notes in Computer Science
Number4232
ISSN0302-9743

Fingerprint Dive into the research topics of 'Probabilistic Kernel Principal Component Analysis Through Time'. Together they form a unique fingerprint.

Cite this