Authors: Stefan Wagner, Gabriel Kronberger,Andreas Beham, Stephan Winkler, Michael Affenzeller
The definition of a generic algorithm model forrepresenting arbitrary heuristic optimization algorithmsis one of the most challenging tasks when developingheuristic optimization software systems. As a highdegree of flexibility and a large amount of reusablecode are requirements that are hard to fulfill together,existing frameworks often lack of either of them to acertain extent. To overcome these difficulties theauthors present a generic algorithm model not onlycapable of representing heuristic optimization but thatcan be used for modeling arbitrary algorithms. Thismodel can be used as a meta-model for heuristicoptimization algorithms, enabling users to representcustom algorithms in a flexible way by still providing abroad spectrum of reusable algorithm building blocks.