Classification of heuristic techniques for performance comparisons

Abstract

In this paper we proposed an enhanced classification of heuristics and developed a unified heuristic classifying scheme and templates for future researchers to conduct a fair heuristic performance. The process involved a thorough diagnostic of the proposed heuristic classification schemes in order to find unique characteristics that can assist us to distinguish each heuristic. We started by carrying out a survey of the heuristic classification schemes and templates previously proposed in the literature. We discovered that most of heuristic classifying schemes and templates proposed in the literature looked at individual heuristic class without integrating together all the classes of heuristics, thus justifying the need for a unified heuristic classifying scheme and templates. We also discovered some common interlinked and interrelated templates of classifying heuristics within which sub-templates were identified. The major challenge in classifying heuristics is basically on discovering unique features (of these heuristics and their variants) that are capable to unlock the element of heuristics belonging to the classifying templates.