Clustering Search with Estimation of Distribution

Abstract

Clustering Search (CS) is a hybrid optimization method which assists the discovery of promising search areas by dividing the search space. The search process is guided by a grouping mechanism that indicates where to perform local search. This paper proposes the use of estimation distribution to assist both the clustering and local search processes in order to reduce the computational effort to optimize combinatorial problems. Computational experiments and theoretical analyzes are used to validate the proposal.