Social Spider and the Prey Search Method for Global Optimization in Hyper Dimensional Search Space

Finding an efficient search strategy to solving complex or difficult problems has been a subject of interest across multiple disciplines particularly in computer science and engineering. In recent times, the applications of metaheuristic algorithms based on evolutionary computation and swarm-based intelligence have demonstrated outstanding performances in search for global optimal solution for any optimization problem. A metaheuristic search algorithm applied in hyper dimensional search space, to find a global optimal solution do not specifically require information about the nature and complexity of the problem because of ease of adapting to some maximum or minimal parameters/constraints of the problem space. In this paper, we present an enhancement to the relatively new developed algorithm called social spider prey (SSP) algorithm for global optimization problem. This algorithm mimics the foraging behavior of social spiders in capturing prey(s) on the social web. The weight of a prey which stimulates the spider's web and cause vibration is depicted and modelled in SSP to enhance the searching strategy of the algorithm particularly, in a hyperdimensional search space. Thus, for improved global optimization algorithm such as SSP to stand the test of time, it is imperative to have it tested on proven benchmarked test functions which is achieved in this paper. A computational experiment was carried out to ascertain the performance of SSP in dealing with complex optimization problems, and the results were discussed. SSP demonstrated outstanding global optimization performance as shown in all the results in all the test functions converging nearly at the global optimum value 0. This study shows the prospects of this relatively new SSP algorithm to solving complex optimization problems.

Automated summary

This paper proposes an enhancement to the SSP algorithm, which mimics the foraging behavior of social spiders, to improve the algorithm's searching strategy. Experimental results demonstrate that SSP performs outstandingly in dealing with complex optimization problems, indicating its prospects for solving such problems.