Mathematics Colloquia and Seminars

Return to Colloquia & Seminar listing

PhD Exit Seminar: Performance Analysis of Spectral Clustering

PDE and Applied Math Seminar

Speaker: Shaofeng Deng
Location: ZOOM
Start time: Tue, Jul 5 2022, 2:00PM

Spectral clustering is a family of graph partitioning methods. Recently, spectral clustering methods have attracted more and more attention given their promising performance in data clustering, simplicity in implementation and efficiency in solving large scale problems.

In this talk, I will discuss the performance of spectral clustering as a graph partitioning tool. I will show, through analyzing the spectral properties of the graph Laplacians, that spectral clustering achieves good results in terms of partitioning both deterministic graphs and random graphs.