Algorithms for independent components analysis and higher order statistics

Daniel D. Lee, Uri Rokni, Haim Sompolinsky

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

A latent variable generative model with finite noise is used to decribe several different algorithms for Independent Components Analysis (ICA). In particular, the Fixed Point ICA algorithm is shown to be equivalent to the Expectation-Maximization algorithm for maximum likelihood under certain constraints, allowing the conditions for global convergence to be elucidated. The algorithms can also be explained by their generic behavior near a singular point where the size of the optimal generative bases vanishes. An expansion of the likelihood about this singular point indicates the role of higher order correlations in determining the features discovered by ICA. The application and convergence of these algorithms are demonstrated on a simple illustrative example.

Original languageEnglish
Title of host publicationAdvances in Neural Information Processing Systems 12 - Proceedings of the 1999 Conference, NIPS 1999
PublisherNeural information processing systems foundation
Pages491-497
Number of pages7
ISBN (Print)0262194503, 9780262194501
StatePublished - 2000
Event13th Annual Neural Information Processing Systems Conference, NIPS 1999 - Denver, CO, United States
Duration: 29 Nov 19994 Dec 1999

Publication series

NameAdvances in Neural Information Processing Systems
ISSN (Print)1049-5258

Conference

Conference13th Annual Neural Information Processing Systems Conference, NIPS 1999
Country/TerritoryUnited States
CityDenver, CO
Period29/11/994/12/99

Fingerprint

Dive into the research topics of 'Algorithms for independent components analysis and higher order statistics'. Together they form a unique fingerprint.

Cite this