Which next iteration is updated yielding an attraction

Which emulates the social behavior of bird flocking and fish schooling,  Artificial Bee Colony (ABC), which is based on the cooperative behavior of bee colonies, Firefly Algorithm (FA) which mimics the mating behavior of firefly insects, and Cuckoo Search (CS), which draws inspiration from the cuckoo bird lifestyle.

Although most of these methods are widely used to solving complex optimization problems, they are known to suffer from some serious flaws, such as premature convergence and the difficulty to overcome local optima, which prevent them from finding optimal solutions. The cause of such issues is usually related to the operators used to modify each individual’s position.

In the case of PSO, for example, the position of each search agent for the next iteration is updated yielding an attraction towards the best particle position seen so-far, while in the case of ABC, positions are updated with respect of some other randomly chosen individuals. As the algorithm evolves, those behaviors allow the entire population to, either rapidly concentrates around the current best particle or to diverge without control, which in return favors the premature convergence or a misbalance between exploration and exploitation respectively. In addition, most state of the art swarm algorithms only model.

Individual entities that perform virtually have the same behavior. Under such circumstances, the possibility of adding new and selective operators based on individual unique characteristics that could improve several important algorithm characteristics such as population diversity and searching capabilities. While it is true that a wide range of organisms living in aggregations shows distinctive cooperative behaviors, this is not true for every single animal species living in social units. In contrast to the popular hypothesis that social behavior is based on mutual benefits for the entire population, the widely accepted selfish herd theory proposed by William D. Hamilton in 1971 illustrates that actions among individuals within aggregations (referred as herds) exhibit an unusual degree of selfishness, particularly when members of such aggregations are endangered by the presence of predators (Hamilton, 1971). In fact, the selfish herd theory establishes that decisions made by any member of such herds do not only benefit the individual itself but also, in exchange, there are usually some negative repercussions for other members on said aggregation. In this paper, a novel swarm optimization algorithm called Selfish Herd Optimizer (SHO) is proposed for solving optimization problems.