報告題目:Estimating the number of significant components in high-dimensional PCA
報告人:潘光明 教授 新加坡南洋理工大學
報告時間:2021年6月14日 9:00至10:00
報告地點:騰訊會議号:906 190 795
校内聯系人:丁雪 dingxue83@jlu.edu.cn
報告摘要:We propose an information criteria to estimate the number of significant components in high-dimensional principal component analysis(PCA). The information criteria is based on the ratio of explained variance and eigenvalue ratios. We show consistency of the estimator in general cases by random matrix theory. We compare its performance with AIC, BIC and some other existing methods for estimating the number of significant components in terms of both theoretical aspects and simulations. An example about the stocks in S&P500 is also reported.
報告人簡介: 潘光明,新加坡南洋理工大學教授。2005年7月博士畢業于中國科學技術大學,自2008年以來,在新加坡南洋理工大學工作。研究領域包括高維統計推斷、随機矩陣理論、多元統計、應用概率等,至今在統計學和概率論的頂級雜志,如: Annals of Statistics, Journal of American Statistical Association, Journal of Royal Statistical Society(B), 《Annals of Probability》、《Annals of Applied Probability》、《Bernoulli》等上發表論文50餘篇。現為國際統計學會會員(Elected Member of International Statistical Institute)。擔任《Random Matrices: Theory and Applications》雜志編委。