PLS: a new statistical insight through the prism of orthogonal polynomials

Mélanie Blazère (1) , Fabrice Gamboa (1) , Jean-Michel Loubes (1)

1 IMT - Institut de Mathématiques de Toulouse UMR5219

Partial Least Square (PLS) is a dimension reduction method used to remove multicollinearities in a regression model. However contrary to Principal Components Analysis (PCA) the PLS components are also choosen to be optimal for predicting the response $Y$. In this paper we provide a new and explicit formula for the residuals. We show that the residuals are completely determined by the spectrum of the design matrix and by the noise on the observations. Because few are known on the behaviour of the PLS components we also investigate their statistical properties in a regression context. New results on regression and prediction error for PLS are stated under the assumption of a low variance of the noise.

Mots clés: Partial Least Square; multivariate regression; multicollinearity; dimension reduction; constrainsted least square; orthogonal polynomials; prediction error

Domaines: Statistiques [math.ST]; Théorie [stat.TH]

Mélanie Blazère, Fabrice Gamboa, Jean-Michel Loubes. PLS: a new statistical insight through the prism of orthogonal polynomials. 2014. ⟨hal-00995049⟩