Many Sparse Cuts via Higher Eigenvalues

TL;DR

This paper generalizes Cheeger's inequality to find multiple sparse cuts in a graph using higher eigenvalues, providing a polynomial-time algorithm with optimal bounds for clustering and expansion.

Contribution

It introduces a spectral method to find multiple sparse cuts based on higher eigenvalues, extending Cheeger's inequality and solving a natural clustering problem.

Findings

Existence of $ck$ disjoint subsets with bounded expansion related to higher eigenvalues

Polynomial-time algorithm combining spectral projection and randomized rounding

Optimal bounds for small set expansion and clustering problems

Abstract

Cheeger's fundamental inequality states that any edge-weighted graph has a vertex subset $S$ such that its expansion (a.k.a. conductance) is bounded as follows: \[ \phi(S) \defeq \frac{w(S,\bar{S})}{\min \set{w(S), w(\bar{S})}} \leq 2\sqrt{\lambda_2} \] where $w$ is the total edge weight of a subset or a cut and $\lambda_2$ is the second smallest eigenvalue of the normalized Laplacian of the graph. Here we prove the following natural generalization: for any integer $k \in [n]$, there exist $ck$ disjoint subsets $S_1, ..., S_{ck}$, such that \[ \max_i \phi(S_i) \leq C \sqrt{\lambda_{k} \log k} \] where $\lambda_i$ is the $i^{th}$ smallest eigenvalue of the normalized Laplacian and $c<1,C>0$ are suitable absolute constants. Our proof is via a polynomial-time algorithm to find such subsets, consisting of a spectral projection and a randomized rounding. As a consequence, we get the same upper bound for the small set expansion problem, namely for any $k$, there is a subset $S$ whose weight is at most a $\bigO(1/k)$ fraction of the total weight and $\phi(S) \le C \sqrt{\lambda_k \log k}$. Both results are the best possible up to constant factors. The underlying algorithmic problem, namely finding $k$ subsets such that the maximum expansion is minimized, besides extending sparse cuts to more than one subset, appears to be a natural clustering problem in its own right.