SFB 303 Discussion Paper No. A-428

Author: Engel, Joachim, and Theo Gasser
Title: A Minimax Result for a Class of Nonparametric Density Estimators
Abstract: A class of kernel type density estimators with locally varying bandwidth is introduced. This class contains the fixed bandwidth estimator,the nearest neighbor estimator, the penaliyed maximum likelihood estimator and a variance stabilizing estimator. While based on asymptotic mean integrated square error (AMISE), there is no uniformly optimal method, the fixed bandwidth estimator is minimax optimal.
Keywords: Nonparametric Density Estimation, Kernel Methods, Nearest Neighbor Estimator, Penalized Maximum Likelihood Estimator, Variance Stabilizing, Minimax Optimality
JEL-Classification-Number: C14, C13, C44
Creation-Date: January 1994
Unfortunately this paper is not available online. Please contact us to order a hardcopy.

SFB 303 Homepage