Negatively Correlated Cooperative Search

Cooperation between individuals is crucial for a population-based meta-heuristic search method to be successful. This paper presents a new population-based search method, namely Negatively Correlated Search (NCS), which essentially maintains multiple individual search processes in parallel. In NCS, the search behaviors of individual search processes are modeled as probability distributions. Negatively correlated search behaviors are explicitly promoted by encouraging difference among the probability distributions (search behaviors). By this means, individual search processes share information and cooperate with each other to search diverse regions of a search space, and makes NCS a promising method for multimodal problems. The cooperation scheme of NCS could also be regarded as a novel diversity preservation scheme that, different from other existing schemes, directly promotes diversity at the level of search behaviors rather than merely trying to maintain diversity among candidate solutions. Empirical studies on a set of multimodal continuous optimization problems showed that NCS is competitive to 6 well-established meta-heuristic search methods in the sense that NCS achieved better solutions on majority of the tested problems.