Random Weighting Estimation Of Kernel Density

Shesheng Gao,Yongmin Zhong

DOI: https://doi.org/10.1016/j.jspi.2010.02.009

IF: 1.095

2010-01-01

Journal of Statistical Planning and Inference

Abstract:Let X-1,X-2,...,X-n be independent and identically distributed random variables with common probability density function f(x). The kernel density estimation of f(x) can be defined as f(n)(x)= (1/nh(n)) Sigma L, K((x-X-i)/h(n)), where K(u) is a kernel function and h(n) > 0 is a series of positive constants that satisfy lim(n ->infinity)h(n) =0. A theory is established to approximate kernel density estimation fn(x) by using random weighting estimation (H) over cap (n)(x) of f(x). Under certain conditions, it rigorously proves that root nh(n)(H) over cap (n)(x)-f(n)(x)) and root nh(n)(f(n)(x)-f(x)) have the same limiting distribution for any random series X-1,X-2,..., X-n. (C) 2010 Elsevier B.V. All rights reserved.