Fatti, L. Paul; Hawkins, Douglas M. Variable selection in heteroscedastic discriminant analysis. (English) Zbl 0594.62065 J. Am. Stat. Assoc. 81, 494-500 (1986). Summary: The likelihood ratio test statistic for the identity in means and covariance matrices of k normal populations has a well-known step-down decomposition measuring the contribution of each component of the vector observation. This decomposition in turn gives rise to three components testing the residual homoscedasticity of each variable, the parallelism of its regression on its predecessors, and the identity of location. A variety of uses of this decomposition in selecting variables is proposed. Cited in 4 Documents MSC: 62H15 Hypothesis testing in multivariate analysis 62H30 Classification and discrimination; cluster analysis (statistical aspects) Keywords:variable selection; heteroscedastic discriminant analysis; factorization of the likelihood ratio statistic; combining test; statistics; means; covariance matrices; normal populations; step-down decomposition; homoscedasticity; location PDFBibTeX XMLCite \textit{L. P. Fatti} and \textit{D. M. Hawkins}, J. Am. Stat. Assoc. 81, 494--500 (1986; Zbl 0594.62065) Full Text: DOI