Optimization Inspired on Herd Immunity Applied to Non-Hierarchical Grouping of Objects

Characterized as one of the most important operations related to data analysis, non-hierarchical grouping consists of, even without having any information about elements be classified, establish upon a finite collection objects, partitioning items that constitute it into subsets or groups intersecting, so are part certain group more similar each other than belong distinct group. In this context, study proposes application meta-heuristic inspired by herd immunity determination and compares results obtained method with answers provided four strategies, described in literature. particular, resulting arrangements classification 33 benchmark collections, performed suggested algorithm, metaheuristic particle swarm, genetic K-means algorithm thermal annealing process, were compared under perspective 10 different evaluation measures, indicating partitions established may, respects, favorable classifications clustering methods.