Enhanced Tunicate Swarm Algorithm for Big Data Optimization

Abstract

Today, with the increasing use of technology tools in daily life, big data has gained even more importance. In recent years, many methods have been used to interpret big data. One of them is metaheuristic algorithms. Meta-heuristic methods, which have been used by very few researchers yet, have become increasingly common. In this study, Tunicate Swarm Algorithm (TSA), which has been newly developed in recent years, was chosen to solve big data optimization problems. The Enhanced TSA (ETSA) was obtained by first developing the swarm action of the TSA. In order to show the achievements of TSA and ETSA, various classical benchmark functions were determined from the literature. The success of ETSA has been proven on these benchmark functions. Then, the successes of TSA and ETSA are shown in detail on big datasets containing six different EEG signal data, with five different population sizes (10, 20, 30, 50, 100) and three different stopping criteria (300, 500, 1000). The results were compared with the Jaya, SOA, and SMA algorithms selected from the literature, and the success of ETSA was determined. The results show that ETSA has sufficient success in solving big data optimization problems and continuous optimization problems.