Fast principal component analysis and data whitening algorithms

Messaoud Thameri*, Abla Kammoun, Karim Abed-Meraim, Adel Belouchrani

*Corresponding author for this work

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

5 Scopus citations

Abstract

In this paper, we propose an adaptive implementation of a fast-convergent algorithm for principal component extraction. Our approach consists of first estimating a basis of the principal subspace through the use of OPAST algorithm. The obtained basis is then fed to a second process where at each iteration one or several Givens transformations are applied to estimate the principal components. Later on, the proposed PCA algorithm is used to derive a fast data whitening solution that overcomes the existing ones of similar complexity order. Simulation results support the high performance of our algorithms in terms of accuracy and speed of convergence.

Original languageEnglish (US)
Title of host publication7th International Workshop on Systems, Signal Processing and their Applications, WoSSPA 2011
Pages139-142
Number of pages4
DOIs
StatePublished - 2011
Externally publishedYes
Event7th International Workshop on Systems, Signal Processing and their Applications, WoSSPA 2011 - Tipaza, Algeria
Duration: May 9 2011May 11 2011

Other

Other7th International Workshop on Systems, Signal Processing and their Applications, WoSSPA 2011
CountryAlgeria
CityTipaza
Period05/9/1105/11/11

Keywords

  • Adaptive algorithms
  • Data whitening
  • Givens rotation
  • Principal component analysis
  • Principal Subspace tracking

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Cite this