Landscape Surveys Through Metaheuristic Algorithms

Abstract

The use of metaheuristic algorithms is proposed as a general method for searching the string landscape for string models with user-specified physical properties. One such algorithm, simulated annealing, is discussed at length and employed to carry out such surveys. Multiple energy functions are considered and it is shown that the particular energy function must be chosen wisely in order to drive the algorithms toward models with desirable properties. It is shown that a simple random sampling outperforms both simulated annealing and random search, likely due to the landscapes highly irregular structure. The chapter concludes with a brief discussion of the application of evolutionary, particularly genetic, algorithms to such surveys.