基于伴随次序统计量的回归函数核估计的强相合性  

Strong Consistance of Kernel Estimation for RegressionFunction Based on Induced Order Statistics

作　　者：凌能祥[1] 

机构地区：[1]合肥工业大学理学院,合肥230009

出　　处：《应用数学》2004年第3期464-467,共4页Mathematica Applicata

摘　　要：本文基于二维随机样本 {(Xi,Yi) ,i≥ 1 }的伴随次序统计量Y[r,n] ,定义了回归函数的核估计 ,在一定条件下 ,获得了回归函数核估计的强相合性 ,推广了已有文献中的部分结果 .The large sample nature of kernal estimation for Regression function have been considered by many authors,which are based on original sample {(X i,Y i),i≥1}.However,in this paper,based on induced statistics Y [r,n] from {(X i,Y i),i≥1},we defined the kernal estimation of regression function and obtained its strong consistance.

关 键 词：伴随次序统计量 核估计 强相合 回归函数 

分 类 号：O212.1[理学—概率论与数理统计] O211.67[理学—数学]