State-of-the-art solvers for Constraint Satisfaction Problems (CSP), Mixed Integer Programs (MIP), and satisfiability problems (SAT) are usually based on a branch-and-bound algorithm. The question how to split a problem into subproblems
is in the core of any branch-and-bound algorithm. Branching on individual variables is very common in CSP, MIP, and SAT. The rules, however, which variable to choose for branching, differ significantly. In this paper, we present hybrid branching, which combines selection rules from all three fields.
Bitte loggen Sie sich ein, um Zugang zu diesem Inhalt zu erhalten