Approximate Counting in Local Lemma Regimes
By: Ryan L. Mann, Gabriel Waite
Potential Business Impact:
Counts hard-to-count things much faster.
We establish efficient approximate counting algorithms for several natural problems in local lemma regimes. In particular, we consider the probability of intersection of events and the dimension of intersection of subspaces. Our approach is based on the cluster expansion method. We obtain fully polynomial-time approximation schemes for both the probability of intersection and the dimension of intersection for commuting projectors. For general projectors, we provide two algorithms: a fully polynomial-time approximation scheme under a global inclusion-exclusion stability condition, and an efficient affine approximation under a spectral gap assumption. As corollaries of our results, we obtain efficient algorithms for approximating the number of satisfying assignments of conjunctive normal form formulae and the dimension of satisfying subspaces of quantum satisfiability formulae.
Similar Papers
Congested Clique Counting for Local Gibbs Distributions
Distributed, Parallel, and Cluster Computing
Counts graph colorings and other complex problems faster.
From Alternation to FPRAS: Toward a Complexity Classification of Approximate Counting
Computational Complexity
Helps computers count hard things faster.
Fast approximate $\ell$-center clustering in high dimensional spaces
Data Structures and Algorithms
Groups data faster, even with many details.