Academics

Universal Rank Inference via Residual Subsampling with Application to Large Networks

Time:4:00-5:00 pm Dec. 7, 2023

Venue:C548 Shuangqing Complex Building 双清综合楼

Organizer:吴昊,杨帆,姜建平,顾陈琳

Speaker:Han Xiao 韩潇 中国科学技术大学

Abstract

Determining the precise rank is an important problem in many large-scale applications with matrix data exploiting low-rank plus noise models. In this paper, we suggest a universal approach to rank inference via residual subsampling (RIRS) for testing and estimating rank in a wide family of models, including many popularly used network models such as the degree corrected mixed membership model as a special case. Our procedure constructs a test statistic via subsampling entries of the residual matrix after extracting the spiked components. The test statistic converges in distribution to the standard normal under the null hypothesis, and diverges to infinity with asymptotic probability one under the alternative hypothesis. The effectiveness of RIRS procedure is justified theoretically, utilizing the asymptotic expansions of eigenvectors and eigenvalues for large random matrices recently developed. The advantages of the newly suggested procedure are demonstrated through several simulation and real data examples.


Speaker

研究方向:

大维随机矩阵,高维统计推断

个人主页:

https://bs.ustc.edu.cn/chinese/profile-652.html


DATEDecember 7, 2023
SHARE
Related News
    • 0

      Skew Howe duality, limit shapes of Young diagrams and universal fluctuations

      AbstractSchur-Weyl, Howe and skew Howe dualities in representation theory of groups lead to multiplicity-free decompositions of certain spaces into irreducible representations and can be used to introduce probability measures on Young diagrams that parameterize irreducible representations. It is interesting to study the behavior of such measures in the limit, when groups become infinite or infi...

    • 1

      Sampling Strategies in Sparse Bayesian Inference

      Mathematics and AI for Imaging Seminars IIOrganizer:Chenglong BaoSpeaker:Yiqiu DongTime:Wednesday, 16:00-17:00Oct. 23, 2024Venue:A04, the 8th FloorShuangqing Complex Building双清综合楼8楼A04Title:Sampling Strategies in Sparse Bayesian InferenceAbstract:Regularization is a common tool in variational inverse problems to impose assumptions on the parameters of the problem. One such assumption...