Speaker：Guangming Pan

Time: 10:45am, Friday, November 22, 2019

Site：EMS A204

Abstract: This talk is about the spiked eigenvalues for high dimensional data with both cross-sectional dependence and dependent sample structure. We illustrate its applications by investigating whether we should implement PCA (factor model) for nonstationary data. A statistic is proposed for distinguishing between unit root models and nonstationary factor model.

Introduction to the Speaker：

Prof. Pan now is a full professor in School of Physical & Mathematical Sciences, Nanyang Technological University. He got his Ph.D degree in mathematical statistics form University of Science and Technology of China in 2005. His research interests include random matrix theory, high dimensional statistics inference, and applications of probability. His has published many articles in top statistics and probability journals like The Annals of Probability, The Annals of Statistics (8), Annals of Applied Probalility, Journal of the American Statistical Association, Journal of the Royal Statistical Society: Series B, Bernoulli and also in econometric journal like Econometric Theory.