Guided operators for a hyper-heuristic genetic algorithm

We have recently introduced a hyper-heuristic genetic algorithm (hyper-GA) with an adaptive length chromosome which aims to evolve an ordering of low-level heuristics so as to find good quality solutions to given problems. The guided mutation and crossover hyper-GA, the focus of this paper, extends that work. The aim of a guided hyper-GA is to make the dynamic removal and insertion of heuristics more efficient, and evolve sequences of heuristics in order to produce promising solutions more effectively. We apply the algorithm to a geographically distributed training staff and course scheduling problem to compare the computational result with the application of other hyper-GAs. In order to show the robustness of hyper-GAs, we apply our methods to a student project presentation scheduling problem in a UK university and compare results with the application of another hyper-heuristic method.