Class

Article

College

College of Science

Department

Mathematics and Statistics Department

Faculty Mentor

Kevin Moon

Presentation Type

Poster Presentation

Abstract

Kernel density estimators are nonparametric techniques used to estimate probability density functions of random variables. Recent work has shown that optimally weighted ensembles of nonparametric estimators can be used to achieve a parametric rate of convergence in various problems, including the problem of estimating information divergence functionals. However, this theory has not yet been adapted to kernel density estimators and is currently being explored. The differences between the information divergence problem and the kernel density estimation case are discussed, as well as various unproven theoretical properties that would imply the existence of a weighted ensemble of kernel density estimators that achieves the parametric rate of convergence. Presentation Time: Thursday, 11 a.m.-12 p.m. Zoom link: https://usu-edu.zoom.us/j/82644853581?pwd=QytsZ1ZFQ3FhUlVzL0NuVHRXRzhYZz09

Location

Logan, UT

Start Date

4-11-2021 12:00 AM

Included in

Life Sciences Commons

Share

COinS
 
Apr 11th, 12:00 AM

Ensemble Kernel Density Estimators

Logan, UT

Kernel density estimators are nonparametric techniques used to estimate probability density functions of random variables. Recent work has shown that optimally weighted ensembles of nonparametric estimators can be used to achieve a parametric rate of convergence in various problems, including the problem of estimating information divergence functionals. However, this theory has not yet been adapted to kernel density estimators and is currently being explored. The differences between the information divergence problem and the kernel density estimation case are discussed, as well as various unproven theoretical properties that would imply the existence of a weighted ensemble of kernel density estimators that achieves the parametric rate of convergence. Presentation Time: Thursday, 11 a.m.-12 p.m. Zoom link: https://usu-edu.zoom.us/j/82644853581?pwd=QytsZ1ZFQ3FhUlVzL0NuVHRXRzhYZz09