Parameter Estimation In Regression Model For Contaminated Data

Parameter estimation in regression model and semi-parametric model for contaminated data is investigated in this paper, which generalizes some known results in the past.The thesis is composed of four parts. In the first part, we make a brief review about contaminated data and some theory about probability and statistics relates to this paper.In the second part, we study the rate of convergence of parameters estimator in linear regression model for contaminated data and only set error having forth moment, the rate of convergence is in accord with Law of iterated logarithm is proved.In part three and four, we explore parameter estimation in semi-parametric model for contaminated data. In part three, we study the rate of convergence of estimations of contamined and parameter in semi-parametric model for contaminated data and prove that rate of convergence is in accord with Law of iterated logarithm. We also discuss asymptotic normality of contaminated parameter and its confidence intervals. In the fourth part, we improve methods of estimation in semi-parametric model for contaminated data and establish their strong consistency under suitable conditions.