m（n）相依样本回归函数改良近邻估计的强相合性

TRONG CONSISTENCY OF THE IMPROVED NEAREST NEIGHBOR ESTIMATE FOR REGRESSION FUNCTION UNDER m(n)-DEPENDENT SAMPLES

研究同分布的ｍ（ｎ）相依样本序列回归函数改良近邻估计的强相合性，同时给出改良近邻估计的强收敛速度。

he authors study the Strong consistency of the improved nearest neighbor estimateof regression function proposed by Cheng Ping and Zhao Lincheng under the condition that {(X_i,X_i)}is a sequence of m(n)-dependent random vectors with common distribution. The strongconvergence rate of the estimate is also obtained.