Multivariate statistical methods: A brief review on their modifications and applications

Article type: Research Article

Authors: Lipovetsky, Stan

Affiliations: Minneapolis, MI, USA | E-mail: stan.lipovetsky@gmail.com

Correspondence: [*] Corresponding author: Minneapolis, MI, USA. E-mail: stan.lipovetsky@gmail.com.

Abstract: The work considers various multivariate statistical techniques in their modifications and applications to management, information systems, economics, decision making, and marketing research problems. The methods include eigenvectors for many-way matrices, dual partial lest squares, modified factor and cluster analyses, and enhanced canonical correlation analysis. These approaches have been applied in numerous real projects and proved to be useful for data analysts, managers, and decision makers in solving practical problems.

Keywords: Multidimensional matrices, three-way SVD and PCA, Dual PLS, factor and cluster analyses, robust and orthonormal canonical correlations

DOI: 10.3233/MAS-220017

Journal: Model Assisted Statistics and Applications, vol. 17, no. 2, pp. 145-147, 2022