The adaptive crossover operator for the traveling salesman problem

Crossover operators have a significant influence on the success of genetic algorithms (GAs). However, a single crossover operator may not be suitable for all instances and different search stages. The basic various crossover operators in the context of the traveling salesman problem (TSP) are discussed. Moreover, a learning-based mechanism for crossover selection is mentioned. The results of experimental comparison show that the application of learning based crossover selection mechanism enables to achieve better solutions than adopting only a single crossover.