报 告 人：潘光明 Associate Professor（新加坡南洋理工大学）
报告题目：The limiting law of the largest eigenvalues and largest nonspike eigenvalue of sample covariance matrices with divergent spikes
报告时间：2017年11月22日（周三）下午 4:00
报告地点：静远楼1506报告厅
报告人简介：
潘光明，2005年中国科学技术大学博士毕业，2008年加入新加坡南洋理工大学，主要研究领域为随机矩阵，高维统计推断，多元分析，计量经济。在Annals of Statistics, Journal of Amercian Statistical Association,Journal of Royal Statistical Society: Series B, Annals of Probability, Annalsof Applied Probability 等概率统计顶级期刊上发表多篇文章。
报告摘要：
We consider sample covariance matrices with divergent eigenvalues in the large dimensional case. We develop the asymptotic distribution of the largest eigenvalues. It has distinct features that the asymptotic mean depends on the population spikes as well as non spikes, and the asymptotic variance depends on the population eigenvectors. The Tracy-Widom distribution for the largest nonspike has been established as well.