An agent-based hyper-heuristic approach to combinatorial optimization problems

Abstract

This paper introduces a framework based on multi-agent system for solving problems of combinatorial optimization. The framework allows running various meta-heuristic algorithms simultaneously. By the collaboration of various meta-heuristics, we can achieve better results in more classes of problems. Our hyper-heuristic approach is defined as a high-level search in algorithm space implemented within multi-agent system.