SPEED IMPROVEMENT OF MULTIVARIATE ALGORITHMS BY THE METHOD OF POSTPONED BASIS MATRIX MULTIPLICATION .2. 3-MODE PRINCIPAL COMPONENT ANALYSIS

Compression is one way of making analysis of large data arrays faster. Compression is here defined as the case when a large array X is replaced by a smaller coefficient array C. The coefficients are obtained by least squares fitting to some compression basis. This paper will deal with three-mode arrays. When performing, e.g., three-mode principal component analysis of C the results are comparable but not equal to the results from analyzing X. In this paper we suggest a solution to this problem by rewriting the three-mode PCA algorithm which utilizes C and the compression basis matrices. This has been accomplished by applying a method where speed improvement is achieved by postponing basis matrix calculations in key steps of the three-mode PCA algorithm.