Seminar Details

Seminar Details

Monday

Aug 14

10:30 am

## Topics in Graph Clustering

### Yali Wan

Final Exam

University of Washington

Second, we describe a model-free framework to provide theoretical guarantees for the results of model-based clustering algorithms. The framework is as follows: if $M(G, C)$ fits the data $G$ well, then we shall prove that any other clustering $C_0$ of $G$ that also fits $G$ well will be a small perturbation of $C$. If this holds, then $C$ with model parameters $M(G, C)$ can be said to capture the data structure in a meaningful way. We instantiate this framework by obtaining model-free guarantees for SBM models.