6533b827fe1ef96bd12874dd

RESEARCH PRODUCT

A note on (basic) Principal Components Analysis

Muggeo Vmr

subject

eigendecomposition multivariate analysisSettore SECS-S/01 - Statistica

description

This paper provides some issues, known but somewhat little stressed, on using the conventional covariance or correlation matrix when performing the simple PCA. The paper also proposes a new simple alternative by providing some evidence, via real-data analysis and some simulation experiments, supporting the proposal

http://hdl.handle.net/10447/434258