Allan Sly

Spectral redemption in clustering sparse networks

Significance Spectral algorithms are widely applied to data clustering problems, including finding communities or partitions in graphs and networks. We propose a way of encoding sparse data using a “nonbacktracking” matrix, and show that the corresponding spectral algorithm performs optimally for some popular generative models, including the stochastic block model. This is in contrast with classical spectral algorithms, based on the adjacency matrix, random walk matrix, and graph Laplacian, which perform poorly in the sparse case, failing significantly above a recently discovered phase transition for the detectability of communities. Further support for the method is provided by experiments on real networks as well as by theoretical arguments and analogies from probability theory, statistical physics, and the theory of random matrices. Spectral algorithms are classic approaches to clustering and community detection in networks. However, for sparse networks the standard versions of these algorithms are suboptimal, in some cases completely failing to detect communities even when other algorithms such as belief propagation can do so. Here, we present a class of spectral algorithms based on a nonbacktracking walk on the directed edges of the graph. The spectrum of this operator is much better-behaved than that of the adjacency matrix or other commonly used matrices, maintaining a strong separation between the bulk eigenvalues and the eigenvalues relevant to community structure even in the sparse case. We show that our algorithm is optimal for graphs generated by the stochastic block model, detecting communities all of the way down to the theoretical limit. We also show the spectrum of the nonbacktracking operator for some real-world networks, illustrating its advantages over traditional spectral clustering.

A proof of the block model threshold conjecture

We study a random graph model called the “stochastic block model” in statistics and the “planted partition model” in theoretical computer science. In its simplest form, this is a random graph with two equal-sized classes of vertices, with a within-class edge probability of q and a between-class edge probability of q′.A striking conjecture of Decelle, Krzkala, Moore and Zdeborová [9], based on deep, non-rigorous ideas from statistical physics, gave a precise prediction for the algorithmic threshold of clustering in the sparse planted partition model. In particular, if q=a/n and q′=b/n, s=(a−b)/2 and d=(a+b)/2, then Decelle et al. conjectured that it is possible to efficiently cluster in a way correlated with the true partition if s2>d and impossible if s2Cdlnd for sufficiently large C.In a previous work, we proved that indeed it is information theoretically impossible to cluster if s2 ≤ d and moreover that it is information theoretically impossible to even estimate the model parameters from the graph when s2 < d. Here we prove the rest of the conjecture by providing an efficient algorithm for clustering in a way that is correlated with the true partition when s2>d. A different independent proof of the same result was recently obtained by Massoulié [20].

Computational Transition at the Uniqueness Threshold

The hardcore model is a model of lattice gas systems which has received much attention in statistical physics, probability theory and theoretical computer science. It is the probability distribution over independent sets

Reconstruction of Markov Random Fields from Samples: Some Observations and Algorithms

I

Belief propagation, robust reconstruction and optimal recovery of block models.

of a graph weighted proportionally to

Proof of the Satisfiability Conjecture for Large k

\lambda^{|I|}

Exact thresholds for Ising–Gibbs samplers on general graphs

with fugacity parameter

Asymptotic Learning on Bayesian Social Networks

\lambda

Maximum independent sets on random regular graphs

. We prove that at the uniqueness threshold of the hardcore model on the

Cutoff phenomena for random walks on random regular graphs

d