An Incremental Reseeding Strategy for Clustering

TL;DR

This paper introduces a simple, parallelizable incremental reseeding algorithm for multiway graph partitioning that achieves state-of-the-art accuracy and significantly faster runtimes compared to existing methods.

Contribution

The paper presents a novel incremental reseeding strategy that improves clustering speed and accuracy, with a scalable approach and a coarsen-refine enhancement.

Findings

Achieves state-of-the-art cluster purity on benchmark datasets.

Runs an order of magnitude faster than comparable algorithms.

Further reduces runtime with a coarsen, cluster, and refine approach.

Abstract

In this work we propose a simple and easily parallelizable algorithm for multiway graph partitioning. The algorithm alternates between three basic components: diffusing seed vertices over the graph, thresholding the diffused seeds, and then randomly reseeding the thresholded clusters. We demonstrate experimentally that the proper combination of these ingredients leads to an algorithm that achieves state-of-the-art performance in terms of cluster purity on standard benchmarks datasets. Moreover, the algorithm runs an order of magnitude faster than the other algorithms that achieve comparable results in terms of accuracy. We also describe a coarsen, cluster and refine approach similar to GRACLUS and METIS that removes an additional order of magnitude from the runtime of our algorithm while still maintaining competitive accuracy.