Seminar No.1532 Multiple Change Point Detection for Correlated High-Dimensional Observations via the Largest Eigenvalue

创建时间:  2017年11月15日 00:00  谭福平   浏览次数:   

Title: Multiple Change Point Detection for Correlated High-Dimensional Observations via the Largest Eigenvalue
Reporter: Prof. Guangming Pan  (Nanyang Technological University, Singapore)
Time: 2017-11-15 (Wednesday) 17:00
Place: G507

Abstract: We propose to deal with a mean vector change point detection problem from a new perspective via the largest eigenvalue when the data dimension p is comparable to the sample size n. An optimization approach is proposed to figure out both the unknown number of change points and multiple change point positions simultaneously. Moreover, an adjustment term is introduced to handle sparse signals when the change only appears in few components out of the p dimensions. The computation time is controlled at $O(n^2)$ by adopting a dynamic programming, regardless of the true number of change points $k_0$. Theoretical results are developed and various simulations are conducted to show the effectiveness of our method.

上一条:Seminar No.1551 Essential sign change numbers and minimum ranks of sign patterns

下一条:Seminar No.1510 A Neighborhood-Assisted Hotelling's t-Test for High-Dimensional Means

CopyRight © Shanghai University    沪ICP备09014157   Address : 99 Shangda Road, BaoShan District, Shanghai.(traffic)   Zip Code : 200444   Tel.
Technical Support : Information Technology Office of Shanghai University   Contact Us