报告题目:

Penalized Maximum Tangent Likelihood Estimation and Robust Variable Selection

报告时间: 7月5日下午4：00-5：30
报告地点: 图东环楼304报告厅
主 讲 人: 李少博

职   称: 助理教授

工作单位: 美国堪萨斯大学商学院 

报告人简介:

李少博，博士，现任美国堪萨斯大学商学院助理教授（AssistantProfessorofBusinessAnalytics）。李少博于2018年毕业于美国辛辛那提大学并取得商业分析博士学位和统计学硕士学位，2011年毕业于山东大学（威海）取得应用数学学士学位。李少博的研究方向包含稳健统计模型，高维统计，非参数模型，数据隐私和安全，公司破产风险预测等。李少博的工作论文曾发表在国际顶级期刊Marketing Science等。李少博曾获统计年会JSM TravelAward和多项校级优秀研究和教学奖。

报告摘要

We introduce a new class of mean regression estimators — penalized maximum tangent likelihood estimation — for high-dimensional robust estimation and variable selection. We first explain the key ingredient, a novel robust method called maximum tangent likelihood estimation (MTE). We establish the optimal rate of convergence in the order of for the proposed penalized MTE under high dimensional regression settings. The proposed penalized MTE has a broad spectrum that consists of penalized distance, penalized exponential squared loss, penalized least trimmed square and penalized least square as special cases, and can be regarded as a mixture of minimum Kullback-Leibler distance estimation and minimum distance estimation. We conduct extensive simulation studies and real data analysis to demonstrate the advantages of the penalized MTE.