吉林大学学报(理学版)

• 数学 • 上一篇    下一篇

两类相依样本密度函数核估计的相合性

胡学平   

  1. 安庆师范学院 数学与计算科学学院, 安徽 安庆 246133
  • 收稿日期:2012-10-22 出版日期:2013-11-26 发布日期:2013-11-21
  • 通讯作者: 胡学平 E-mail:hxprob@163.com

Consistency of KernelType Density Estimations ofTwo Sorts of Dependent Samples

HU Xueping   

  1. School of Mathematics and Computational Science, Anqing Normal University, Anqing 246133, Anhui Province, China
  • Received:2012-10-22 Online:2013-11-26 Published:2013-11-21
  • Contact: HU Xueping E-mail:hxprob@163.com

摘要:

设{Xn,n≥1}为同分布的NOD随机序列或严平稳的m相依序列, f(x)为随机变量X1的概率密度函数. 基于样本X1,X2,…,Xn, 利用Fourier变换及NOD列的性质和相关指数不等式, 研究密度函数f(x)的核估计, 在适当的条件下得到了[KG-*4]f(x)核估计的逐点强相合性、 r阶相合性及依概率一致收敛性.

关键词: NOD样本, m相依样本, 核估计, 强相合性, r阶相合性

Abstract:

Let {Xn,n≥1} be an identically distributed NOD random sequence or a strict stationary sequence of m dependent random variables, and f(x) is a probability density function of random variable X1. Based on NOD samples or m dependent samples, under suitable conditions the kernel estimator for density function f(x) was investigated, and the pointwise strong consistency, the consistency in r order mean and uniform consistency in L1 were obtained via Fourier transform, the related properties for NOD random sequences and the exponential inequality.

Key words: NOD samples, m dependent samples, kernel estimator, strong consistency, consistency in r order mean

中图分类号: 

  • O212.2