Understanding the power of Max-SAT resolution through UP-resilience
Abstract
Input: a formula Φ in Conjunctive Normal Form (CNF) Output: the maximum (resp. minimum) number of satisfied (resp. falsified) clauses in Φ over all possible variable assignments Branch & Bound (BnB) for Max-SAT Binary search algorithm which maintains and constantly updates two values : Upper Bound (UB): value of the best known solution Lower Bound (LB): estimation of the best accessible solution Cut: if LB ≥ UB then backtrack
Domains
Artificial Intelligence [cs.AI]
Origin : Files produced by the author(s)