This article provides an introduction to hyperheuristics, the origins of which can be traced back to the 1960s. There was a significant increase in research interest from about 2000. Hyperheuristics can be seen as either
heuristics to choose heuristics or as heuristics to generate heuristics.
In contrast to most uses of metaheuristics, which usually search the solution space directly, hyperheuristics search over a heuristic space. One motivation of the area is that, by operating in this way it might be possible to raise the level of generality of search algorithms so that they can address a broader spectrum of problems (without human adaptation) than is possible today. Another long‐term aim of hyperheuristic research is to build search methodologies that are able to automatically adapt themselves to different problem instances/domains with as little human input as possible.
This article provides access to relevant literature, in addition to providing a brief history of the area and describing the two approaches (selection of heuristics and generating heuristics) mentioned above. It is intended as a brief and initial introduction to the area.