Optimal quantization applied to Sliced Inverse Regression
Statistics Theory
2011-01-13 v2 Statistics Theory
Abstract
In this paper we consider a semiparametric regression model involving a -dimensional quantitative explanatory variable and including a dimension reduction of via an index . In this model, the main goal is to estimate the euclidean parameter and to predict the real response variable conditionally to . Our approach is based on sliced inverse regression (SIR) method and optimal quantization in -norm. We obtain the convergence of the proposed estimators of and of the conditional distribution. Simulation studies show the good numerical behavior of the proposed estimators for finite sample size.
Cite
@article{arxiv.1101.2121,
title = {Optimal quantization applied to Sliced Inverse Regression},
author = {Azaïs Romain and Gégout-Petit Anne and Saracco Jérôme},
journal= {arXiv preprint arXiv:1101.2121},
year = {2011}
}