Practical sequential method for principal component analysis

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal

6 Scopus Citations
View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)107-112
Journal / PublicationNeural Processing Letters
Volume11
Issue number2
Publication statusPublished - 2000

Abstract

When increasing numbers of principal components are extracted by using the sequential method proposed in [1] by Banour and Azimi-Sadjadi, the accumulated extraction error will become dominant and affect the extractions of the remaining principal components. To improve this, we suggest that the initial weight vector for the extraction of the next component should be orthogonal to the eigensubspace spanned by the already extracted weight vectors. Simulation results show that both the convergence and the accuracy of the extraction are improved. Our improved method is also capable of extracting full eigenspace accurately.

Citation Format(s)

Practical sequential method for principal component analysis. / Wong, Arnold Shu-Yan; Wong, Kwok-Wo; Leung, Chi-Sing.

In: Neural Processing Letters, Vol. 11, No. 2, 2000, p. 107-112.

Research output: Journal Publications and Reviews (RGC: 21, 22, 62)21_Publication in refereed journal