A semi-Bayesian method for nonparametric density estimation

被引：9

作者：

de Bruin, R

Salomé, D

Schaafsma, W

机构：

[1] Univ Groningen, Ctr High Performance Comp, NL-9700 AV Groningen, Netherlands

[2] Univ Groningen, Dept Math & Comp Sci, NL-9700 AV Groningen, Netherlands

来源：

COMPUTATIONAL STATISTICS & DATA ANALYSIS | 1999年 / 30卷 / 01期

关键词：

D O I：

10.1016/S0167-9473(98)00089-9

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

If x([1],) ..., x([n]) are the ordered outcomes of an independent random sample from a distribution on the interval [x([0]), x([n+1])] = [a, b] with distribution function F and density function f then 'intrinsic' arguments (Section 2) lead to a specific Bernstein polynomial as 'the' estimate of F-1(p). Numerical inversion and differentiation provides 'the' estimate of f = F' which we study in detail. A comparison is made, extensions are indicated and literature is discussed. (C) 1999 Elsevier Science B.V. All rights reserved.

引用

页码：19 / 30

页数：12

共 20 条

[1] SOME MINIMAX INVARIANT PROCEDURES FOR ESTIMATING A CUMULATIVE DISTRIBUTION FUNCTION [J].