Factor Analysis as Data Matrix Decomposition: A New Approach for Quasi-Sphering in Noisy ICA

Authors:
Steffen Unkel;Nickolay T. Trendafilov
Affiliations:
Department of Mathematics and Statistics, Faculty of Mathematics, Computing & Technology, The Open University, Walton Hall, Milton Keynes, United Kingdom MK7 6AA;Department of Mathematics and Statistics, Faculty of Mathematics, Computing & Technology, The Open University, Walton Hall, Milton Keynes, United Kingdom MK7 6AA
Venue:
ICA '09 Proceedings of the 8th International Conference on Independent Component Analysis and Signal Separation
Year:
2009

Citing 2
Cited 0

Matrix computations (3rd ed.)

Matrix computations (3rd ed.)
Maximum likelihood factor analysis with rank-deficient sample covariance matrices

Journal of Multivariate Analysis

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, a new approach for quasi-sphering in noisy ICA by means of exploratory factor analysis (EFA) is introduced. The EFA model is considered as a novel form of data matrix decomposition. By factoring the data matrix, estimates for all EFA model parameters are obtained simultaneously. After the preprocessing, an existing ICA algorithm can be used to rotate the sphered factor scores towards independence. An application to climate data is presented to illustrate the proposed approach.