Total Information in Multivariate Data from Dual Scaling Perspectives

It is an established matter that the total information in multivariate data is defined as the sum of eigenvalues of the variance-covariance matrix. In this article I challenge this time-honored tradition and look at another definition of the total information in data from a dual scaling perspective. This proposal is a step toward unifying the concept of information for both discrete and continuous variables

Covariance · Covariance matrix · Data Matrix · Dual (grammatical number) · Econometrics · Eigenvalues and eigenvectors · Multidimensional scaling · Multivariate analysis · Multivariate statistics · Perspective (graphical) · Physics · Scaling · Statistics · Variance (accounting) · Artificial Intelligence · Bayesian Modeling and Causal Inference · Computer Science · Mathematics · Philosophy