报告地点:行健楼学术活动室526
Abstract:.We consider four structures of high dimensional time series in terms of factor structure and nonstationary. We propose a novel approach to identifying them. The proposed three-step method includes:
(1) the ratio statistic of empirical eigenvalues;
(2) a projected Augmented Dickey-Fuller Test;
(3) a new unit-root test based on the largest empirical eigenvalues.
专家简介:潘光明,新加坡南洋理工大学教授,博士生导师。2002年硕士毕业于安徽大学,2005年博士毕业于中国科学技术大学,之后在新加坡国立大学、台湾国立中山大学、荷兰埃因霍温科技大学做博士后和学术交流工作;自2008年以来,在新加坡南洋理工大学工作。研究领域包括高维统计推断、随机矩阵理论、多元统计、应用概率等。至今已在Annals ofStatistics、Journal of American Statistical Association、Journal ofRoyal Statistical Society ( B)、Annals of Probability、Annals ofApplied Probability、Bernoulli、IEEE Transactions on Signal Pro-cessing、IEEE Transactions on Information Theory等顶级统计学杂志上发表60余篇学术论文。现为国际统计学会会员(Elected Member ofnternational Statistical Institute),《Random Matrices: Theory andApplications》杂志编委。