Convergence rate for predictive recursion estimation of finite mixtures
[作者] Ryan Martin
[单位] Department of Mathematical Sciences Indiana University–Purdue University Indianapolis
[摘要] Predictive recursion (PR) is a fast stochastic algorithm for nonparametric estimation of mixing distributions in mixture models.
[关键词] Density estimation Kullback–Leibler divergence
Predictive recursion (PR) is a fast stochastic algorithm for nonparametric estimation of mixing distributions in mixture models. It is known that the PR estimates of both the mixing and mixture densities are consistent under fairly mild conditions, but currently very little is known about the rate of convergence. In this note we investigate asymptotic convergence properties of the PR estimate under model mis-specification in the special case of finite mixtures with known support. Tools from stochastic approximation are used to prove that the PR estimates converge at a nearly root-n rate. This result provides some important clues about the choice of weight sequence in the PR algorithm in general.
存档附件原文地址
原文发布时间:2011/7/6
引用本文:
Ryan Martin.Convergence rate for predictive recursion estimation of finite mixtures.http://hftc.firstlight.cn/View.aspx?infoid=1034112&cb=hangxiaojingxg.
发布时间:2011/7/6.检索时间:2024/12/15