报告题目：SIMEX estimation for single-index measurementerror models

报 告 人：李高荣教授（北京工业大学）

报告时间：11月9日（周1）下午3:00-4:00

报告地点：知新楼B-1238

摘要：In this paper, we consider the single-indexmeasurement error model when mismeasured covariate appears in the noparametricpart. Based on the local linear smoother and the estimating equation, wedevelop a  simulation-extrapolation(SIMEX) algorithm to solve this problem. One important feature of the SIMEXestimation  is that there is no need toassume the distribution of the unobservable covariate. The other feature isthat we transform the boundary of a unit ball in $\mathbb{R}^p$ to the interiorof a unit ball in $\mathbb{R}^{p-1}$ by using the constraint $\|\beta\|=1$. Theproposed SIMEX estimator of the index parameter is shown to be asymptoticallynormal under some regularity conditions. We derive the asymptotic bias andvariance of the estimator of the unknown link function. The finite sampleperformance of the proposed estimation is examined via simulation studies. Theproposed method is applied to a real data.

欢迎各位老师和同学参加！