Conditional Kernel Density Estimation For Dependent Data, Mean Shift Algorithm and Predictive Regions in Human Health
- Prof. Mohamed El Machkouri, Associate professor, Applied Mathematics, University of Rouen Normandy
B1 L4 R4102
In this talk, we present theoretical asymptotic results for the nonparametric estimation of the conditional density of a scalar response variable Y given the explanatory X taking values in a finite-dimensional space when the sample of observations is considered as a sequence of dependent random variables.
Overview
Abstract
In this talk, we present theoretical asymptotic results for the nonparametric estimation of the conditional density of a scalar response variable Y given the explanatory X taking values in a finite-dimensional space when the sample of observations is considered as a sequence of dependent random variables. As an application, using the so-called mean-shift algorithm, we provide predictive regions for bioimpedance fat-free mass for anorectic patients. This is a joint work with Pr. Najate Achamrah and Pr. Moise Coëffier (INSERM UMR 1073, Department of Clinical Nutrition of the Rouen University Hospital).
Brief Biography
Mohamed EL MACHKOURI is Associate Professor in applied mathematics at University of Rouen Normandy (France). He is a member of the Mathematical Laboratory Raphaël Salem (LMRS UMR 6085). His main research interests are related to limit theorems for dependent sequences and fields of random variables in probability theory with applications to nonparametric statistics and machine learning.
