A Novel Metaheuristic: Fast Jaguar Algorithm
Shu–Yu Kuo, Ching-Hsuan Wu, Cheng-Chun Chen, Yao–Hsin Chou · 2021 IEEE International Conference on Systems, Man, and Cybernetics (SMC) · 2021
Metaheuristic algorithms play an extremely important role in the computational intelligence and optimization fields. The jaguar algorithm (JA) is a metaheuristic algorithm that has outstanding performances in both exploitation and exploration. This study proposed a novel metaheuristic algorithm, named the fast jaguar algorithm (FJA), which inherits JA’s advantage and significantly enhances its search ability. FJA uses the fast hunting mechanism to identify the trend in the local area effectively, and it only searches for the better side to reduce the computational cost. FJA well uses historical information through the adaptive exploit mechanism, which allows it to efficiently find the better tendency. Then, FJA utilizes the jump mechanism to discover the global optimal solution. FJA comprehensively considers more situations and then makes the right decision to find the optimal solution more precisely. The experiment results demonstrated the robust performance of FJA through function optimization and showed that the efficiency of FJA could outperform JA and other classical metaheuristics in terms of computational cost and stability.