2008•Journal of Jinling Institute of TechnologyRequires access
Asymptotic Normality of Improved Kernel Estimator for Regression Function under α-mixing Samples
Wenfeng Zhang
Open publisher page 0 citations
Abstract
Let{(Xi,Yi),i≥1}be a strictly stationary andα-mixing sample sequence fromi≥1 in Rd×R.Thei mproved kernel esti mator for regressionfunctionm(x)=E(Y|X=x)is defined by m2n(x) =∑ni =1YiI(| Yi| bi) hi-dKx-hiXi∑nj=1hj-dKx-hjXj Under suitable conditions,we prove the asymptotic normality ofm2n(x).