Local Search for Statistical Counting

In this paper, statistical counting is introduced in the context of stochastic local search. From a sample of trajectories by independent local search computations, it is shown that interesting statistical information can be actually extracted about the search space, most notably an unbiased estimate of the number of solutions. Computational results for random #SAT instances are provided.

Authors

• Olivier Bailleux

Keywords

No keywords are indexed for this paper.