Variable importance assessment in sliced inverse regression for variable selection
Résumé
We are interested in treating the relationship between a dependent
variable $y$ and a multivariate covariate $x \in {\R}^p$ in a
semiparametric regression model. Since the purpose of most social,
biological or environmental science research is the explanation, the determination of the
importance of the variables is a major concern. It is a way to
determine which variables are the most important when predicting
$y$. Sliced inverse regression methods allows to reduce the space of the
covariate $x$ by estimating the directions $\beta$ that form an
effective dimension reduction (EDR) space. The aim of this paper is
to propose a computational method based on importance variable measure (only relying on the EDR space) in order to select the most useful variables. The numerical behavior of this new method, implemented in R, is studied on a simulation study. An illustration on a real data is also provided.