Scaling and Probabilistic Smoothing for SAT and MAX-SAT

By Dave Tompkins (joint work with Holger Hoos and Frank Hutter)

In our presentation we will discuss the Dynamic Local Search (DLS) algorithm known as Scaling and Probabilistic (SAPS). We will cover the motivation for developing SAPS, and then demonstrate its' effectiveness on SAT problem instances. We will also provide some insight into why the SAPS algorithm performs so well. We will then discuss the merits of applying SAPS to the weighted and unweighted MAX-SAT problem domain, and provide some empirical results from our MAX-SAT experiments.