报告题目:Sparse Sir: Optimal Rates and Adaptive Estimation
时间:6月18日 (本周二) 下午13:00—14:00
地点:博识楼434
摘要: Sliced inverse regression (SIR) is an innovativeand effective method for sufficient dimension reduction and data visualization.Recently, an impressive range of penalized SIR methods has been proposed to estimatethe central subspace in a sparse fashion. Nonetheless, few of them consideredthe sparse sufficient dimension reduction from a decision-theoretic point of view.To address this issue, we in this paper establish the minimax rates ofconvergence for estimating the sparse SIR directions under various commonlyused loss functions in the literature of sufficient dimension reduction. Wealso discover the possible trade-o_ between statistical guarantee andcomputational performance for sparse SIR. We finally propose an adaptiveestimation scheme for sparse SIR which is computationally tractable and rateoptimal. Numerical studies are carried out to confirm the theoreticalproperties of our proposed methods.
报告人简介:於州,华东师范大学博士,香港浸会大学博士后,美国威斯康辛大学麦迪逊分校以及美国国家统计局高级访问学者。於州老师于2013年和2016年分别被破格晋升为副教授和教授,现为华东师大教授、博士生导师,研究方向为高维数据分析,曾主持两项国家自然科学基金、上海市青年科技启明星计划等项目,获得上海市东方特聘教授称号,获得第十届全国统计科研成果奖二等奖(博士论文类),其博士毕业论文被评为上海市优秀博士毕业论文。目前於州老师担任社会职务有中国现场统计研究会副秘书长、全国工业统计教学研究会常务理事、中国概率统计学会理事、中国现场统计研究会高维数据统计分会常务理事。