English

Robust Hypergraph Clustering via Convex Relaxation of Truncated MLE

Machine Learning 2020-11-19 v3 Information Theory Machine Learning math.IT

Abstract

We study hypergraph clustering in the weighted dd-uniform hypergraph stochastic block model (dd\textsf{-WHSBM}), where each edge consisting of dd nodes from the same community has higher expected weight than the edges consisting of nodes from different communities. We propose a new hypergraph clustering algorithm, called \textsf{CRTMLE}, and provide its performance guarantee under the dd\textsf{-WHSBM} for general parameter regimes. We show that the proposed method achieves the order-wise optimal or the best existing results for approximately balanced community sizes. Moreover, our results settle the first recovery guarantees for growing number of clusters of unbalanced sizes. Involving theoretical analysis and empirical results, we demonstrate the robustness of our algorithm against the unbalancedness of community sizes or the presence of outlier nodes.

Keywords

Cite

@article{arxiv.2003.10038,
  title  = {Robust Hypergraph Clustering via Convex Relaxation of Truncated MLE},
  author = {Jeonghwan Lee and Daesung Kim and Hye Won Chung},
  journal= {arXiv preprint arXiv:2003.10038},
  year   = {2020}
}

Comments

20 pages, 4 figure