Performance evaluation of nature inspired meta-heuristic algorithms using rosenbrock, rastrigin and sphere test function for optimization

Abstract

Due to a large pool of optimization algorithms available, it is a daunting task for a researcher to select the best algorithm for their optimization problem. Nature inspired meta-heuristic algorithms have emerged and are very popular among researchers for their research. The meta-heuristic algorithms are used, for finding and generating partial search algorithm that is supposed to provide a sufficiently good solution to any optimization problem, especially with imperfect information. In this research paper, we have chosen two widely researched and rapidly developing algorithms namely Firefly algorithms (FA) and Artificial Bee Colony (ABC) Algorithms to test on different test functions. Therefore, in this research paper the efficiency of Artificial Bee Colony and Firefly algorithms are rigorously tested on three most used test function for optimization namely Rosenbrock test function, Rastrigin test function and Sphere test function. The best cost for both the algorithms on, all three test functions are presented and compared to find which algorithm is the best along with the test function for finding an optimal solution to meta-heuristic implementation.