Vol.:(0123456789) Artificial Intelligence Review https://doi.org/10.1007/s10462-020-09931-5 1 3 Improved social spider algorithm for large scale optimization Emine Baş 1  · Erkan Ülker 2 © Springer Nature B.V. 2020 Abstract Heuristic algorithms can give optimal solutions for low, middle, and large scale optimi- zation problems in an acceptable time. The social spider algorithm (SSA) is one of the recent meta-heuristic algorithms that imitate the behaviors of the spider to perform global optimization. The original study of this algorithm was proposed to solve low scale con- tinuous problems, and it is not be solved to middle and large scale continuous problems. In this paper, we have improved the SSA and have solved middle and large scale continu- ous problems, too. By adding two new techniques to the original SSA, the performance of the original SSA has been improved and it is named as an improved SSA (ISSA). In this paper, various unimodal and multimodal standard benchmark functions for low, middle, and large-scale optimization are studied for displaying the performance of ISSA. ISSA’s performance is also compared with the well-known and new evolutionary methods in the literature. Test results show that ISSA displays good performance and can be used as an alternative method for large scale optimization. Keywords Heuristic methods · Optimization · Social spider algorithm · Large-scale dimension 1 Introductıon For the last 20 years, swarm intelligence-based algorithms attract attention. The term of a swarm expresses a community that consists of individuals who are in communication with each other. Swarm intelligence-based algorithms consider social animal and insect behav- iors for solving problems. These algorithms imitate behaviors of ant, bee, bacteria, but- terfly, etc. Thus, the problems which seem hard can be solved (Parpinelli and Lopes 2011; Yu and Li 2015). A computational optimization methodology involves finding feasible solutions from a finite set of solutions and identifying only the optimal solution(s). Swarm * Emine Baş emineozcan@selcuk.edu.tr Erkan Ülker eulker@ktun.edu.tr 1 Kulu Vocational School, Selçuk University, 42075 Konya, Turkey 2 Department of Computer Engineering, Faculty of Engineering and Nature Sciences, Konya Technical University, 42075 Konya, Turkey