Machine learning at the service of meta-heuristics for solving combinatorial optimization problems: A state-of-the-art

In recent years, there has been a growing research interest in integrating machine learning techniques into meta-heuristics for solving combinatorial optimization problems. This integration aims to lead toward an efficient, effective, and robust search improve their performance terms of solution quality, convergence rate, robustness. Since various methods with different purposes have developed, is need review the advances using meta-heuristics. To best our knowledge, literature deprived having comprehensive yet technical review. fill this gap, paper provides such on use design elements including algorithm selection, fitness evaluation, initialization, evolution, parameter setting, cooperation. First, we describe key concepts preliminaries each these ways integration. Then, way are reviewed classified based proposed unified taxonomy. Finally, provide discussion advantages, limitations, requirements, challenges implementing ways, followed by promising future directions.