污染数据回归分析中参数的最小一乘估计

Least Absolute Deviation Estimator of Parameter in Regression Analysis for Contamination Data

考虑简单回归模型：其中 ．假设ｙ１， ｙ２，… ；ｙｎ受到另一独立同分布随机变量序列的污染；我们仅能观察到污染数据为未知的污染参数．本文用污染数据给出了的最小一乘估计，并证明了它的渐近正态性和相合性．

Consider the simple regression model , where Eei= 0, Ee2ic = Assume that y1, y2, yn are contaminated by another i.i.d. random variable variable sequence and only the contaminated data yi = are observable, where is unknown contaminated parameter. In this paper we give LADE of by contamination data and establish its asymptotic normality and consistency.