Multiway Spectral Partitioning and Higher-Order Cheeger Inequalities

• Published in: Journal of the ACM Vol. 61; no. 6; pp. 1 - 30
• Main Authors: Lee, James R., Gharan, Shayan Oveis, Trevisan, Luca
• Format: Journal Article
• Language: English
• Published: New York Association for Computing Machinery 01.11.2014
• Subjects: Algorithms; Approximation; Constants; Eigenvalues; Graph theory; Graphs; Inequalities; Laplace transforms; Matrix; Optimization; Spectra; Studies
• ISSN: 0004-5411; 1557-735X
• DOI: 10.1145/2665063

A basic fact in spectral graph theory is that the number of connected components in an undirected graph is equal to the multiplicity of the eigenvalue zero in the Laplacian matrix of the graph. In particular, the graph is disconnected if and only if there are at least two eigenvalues equal to zero. Cheeger's inequality and its variants provide an approximate version of the latter fact; they state that a graph has a sparse cut if and only if there are at least two eigenvalues that are close to zero. It has been conjectured that an analogous characterization holds for higher multiplicities: There are k eigenvalues close to zero if and only if the vertex set can be partitioned into k subsets, each defining a sparse cut. We resolve this conjecture positively. Our result provides a theoretical justification for clustering algorithms that use the bottom k eigenvectors to embed the vertices into R k , and then apply geometric considerations to the embedding. We also show that these techniques yield a nearly optimal quantitative connection between the expansion of sets of size ≈ n / k and λ k , the k th smallest eigenvalue of the normalized Laplacian, where n is the number of vertices. In particular, we show that in every graph there are at least k /2 disjoint sets (one of which will have size at most 2 n / k ), each having expansion at most O (√λ k log k ). Louis, Raghavendra, Tetali, and Vempala have independently proved a slightly weaker version of this last result. The √log k bound is tight, up to constant factors, for the “noisy hypercube” graphs.