the 23rd Benelux Conference on Artificial Intelligence (BNAIC11)
the 23rd Benelux Conference on Artificial Intelligence (BNAIC11) location:Gent, Belgium date:3-4 November 2011
Selection hyper-heuristics concentrate on using the strength of multiple low-level search mechanisms for solving instances from various problem domains. A traditional selection hyper-heuristic is composed of 1) a heuristic selection mechanism for choosing heuristics at each decision step and 2) a move acceptance strategy for deciding about whether or not to use the explored solutions by the selected heuristics. These mechanisms work together in a problem-independent manner to raise the level of generality on the one hand and to ease their applicability on the other hand. The present study provides a new selection hyper-heuristic equipped with various adaptive features.