A Scalable MapReduce-enabled Glowworm Swarm Optimization Approach for High Dimensional Multimodal Functions_sys_v5

Highly multimodal function optimization is similar to many other optimization problems requiring many iterations and large number of function evaluations. Glowworm Swarm Optimization (GSO) is one of the common swarm intelligence algorithms, where GSO has the ability to optimize multimodal functions efficiently. Locating the peaks of a high-dimensional multimodal function requires a large population, which is considered time consuming when sequential algorithms are used. Moreover, increasing the number of dimensions of a multimodal function usually increases the number of peaks exponentially. Therefore, a parallelization of the GSO algorithm is necessary to reduce the long execution time for capturing the peaks. In this paper, a parallel MapReduce-based GSO algorithm is proposed to speedup the GSO optimization process. We argue that GSO can be formulated based on the MapReduce parallel programming model quite naturally. In addition, we use higher dimensional multimodal benchmark functions for evaluating the proposed algorithm. The experimental results show that the proposed algorithm is appropriate for optimizing difficult multimodal functions with higher dimensions and achieving high peak capture rates. Furthermore, a scalability analysis shows that the proposed algorithm scales very well with increasing swarm sizes, and scales very close to the linear speedup while maintain high peak capture rates. In addition, an overhead of the Hadoop infrastructure is investigated to find if there is any relationship between the overhead, the swarm size, and number of nodes used.