报告人： Associate Prof. Xuejun Jiang(蒋学军), Department of Statistics and Data Science, Southern University of Science and Technology, Shenzhen, China
报告摘要: In this paper we propose an approximately oracle general likelihood ratio test(HOGLR) to test linear hypotheses in high dimensional linear models. To deal with linear hypotheses test, we first investigate the difficulty in using the ordinary least square estimation and then propose the projection least square estimation (PLSE) to surmount it. The PLSE is essentially an ordinary LSE based on the true feature space. Based on the PLSE, we construct a generalized likelihood ration statistics which compares the negative Gaussian quasi-loglikelihood functions under the null and the alternatives. We show that the generalized likelihood rato test enjoys Wilks phenomenon under null hypothesis, and under the alternatives, it asymptotically follows non-central χ2 distribution with the same freedom as the limiting null distribution. Simulations are conducted to examine the finite sample performance of the proposed test. Empirical analysis of a real data example is used to illustrate the use of our proposed testing procedure.
会议时间：2020/7/3 14:00-16:30
实际时间：14:30-16：00
点击链接入会，或添加至会议列表：
https://meeting.tencent.com/s/ZS6b5Yit6KoW
会议 ID：628 826 939
会议密码：703703