Finding Sparse Cuts Locally Using Evolving Sets

TL;DR

This paper introduces a new local graph partitioning algorithm based on the volume-biased evolving set process, which finds sparse cuts more efficiently and with better approximation guarantees than previous methods.

Contribution

The paper presents a randomized local partitioning algorithm that improves approximation guarantees and reduces complexity ratios compared to prior algorithms, using a novel evolving set process approach.

Findings

Achieves conductance approximation of O(φ^{1/2} log^{1/2} n) for at least half of the starting vertices in set A.

Expected ratio of computational complexity to output volume is O(φ^{-1/2} polylog(n)).

Provides a fast algorithm for finding balanced cuts with improved efficiency and conductance guarantees.

Abstract

A {\em local graph partitioning algorithm} finds a set of vertices with small conductance (i.e. a sparse cut) by adaptively exploring part of a large graph $G$, starting from a specified vertex. For the algorithm to be local, its complexity must be bounded in terms of the size of the set that it outputs, with at most a weak dependence on the number $n$ of vertices in $G$. Previous local partitioning algorithms find sparse cuts using random walks and personalized PageRank. In this paper, we introduce a randomized local partitioning algorithm that finds a sparse cut by simulating the {\em volume-biased evolving set process}, which is a Markov chain on sets of vertices. We prove that for any set of vertices $A$ that has conductance at most $\phi$, for at least half of the starting vertices in $A$ our algorithm will output (with probability at least half), a set of conductance $O(\phi^{1/2} \log^{1/2} n)$. We prove that for a given run of the algorithm, the expected ratio between its computational complexity and the volume of the set that it outputs is $O(\phi^{-1/2} polylog(n))$. In comparison, the best previous local partitioning algorithm, due to Andersen, Chung, and Lang, has the same approximation guarantee, but a larger ratio of $O(\phi^{-1} polylog(n))$ between the complexity and output volume. Using our local partitioning algorithm as a subroutine, we construct a fast algorithm for finding balanced cuts. Given a fixed value of $\phi$, the resulting algorithm has complexity $O((m+n\phi^{-1/2}) polylog(n))$ and returns a cut with conductance $O(\phi^{1/2} \log^{1/2} n)$ and volume at least $v_{\phi}/2$, where $v_{\phi}$ is the largest volume of any set with conductance at most $\phi$.