Kernel Generalized Canonical Correlation Analysis

Arthur Tenenhaus (1)

1 E3S - Supélec Sciences des Systèmes

A classical problem in statistics is to study relationships between several blocks of variables. The goal is to find variables of one block directly related to variables of other blocks. The Regularized Generalized Canonical Correlation Analysis (RGCCA) is a very attractive framework to study such a kind of relationships between blocks. However, RGCCA captures linear relations between blocks and to assess nonlinear relations we propose a kernel extension of RGCCA.

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

Arthur Tenenhaus. Kernel Generalized Canonical Correlation Analysis. JdS'10, May 2010, Marseille, France. CD-ROM Proceedings (6 p.). ⟨hal-00553602⟩