COMPUTER AIDED SELECTION IN DESIGN PROCESSES WITH MULTIVARIATE STATISTICS

: Principal Component Analysis (known briefly as PCA) is a multivariate statistical technique for simplifying a cloud data set [1, 2]. Based on this procedure the observable possibly correlated properties are reported into a few uncorrelated “attributes”; in other words it is considered a transformation from a space into a subspace such that the retained variance of the original cloud is “maximal” by this new representation [3, 4]. In a previous paper [7] we applied the so called the Jöreskog' technique used for the dimensional reduction in a bivariate subspace. The goal of this paper is to apply the Pearson method of PCA, for the same family of materials, using the XLSTAT 2009 software; new relevant factors are obtained, and the results are comparable.