Multimeme Algorithms Using Fuzzy Logic Based Memes For Protein Structure Prediction

Abstract

In this chapter we extend our previous studies on the self-adaptation of local searchers within a Memetic Algorithm. Self-adaptation allows the MA to learn which local searcher to use during search. In particular, we extend our results in tikya[12], where memes were instantiated as Fuzzy-Logic based local searchers, and we show that our Multimeme algorithms are capable of producing new optimum solutions to instances of the Protein Structure Prediction Problem in the HP-model.