Comparison between Fast Evolutionary Programming and Artificial Bee Colony Algorithm on Numeric Function Optimization Problems

The Evolutionary and Swarm Intelligence algorithms are two recently introduced population based meta-heuristic algorithms that have been successfully employed to numerous scientific and engineering problems. In this paper, we have selected two recent and representative algorithms one from the evolutionary algorithm family, the other from the swarm intelligence family and compared their performance on high dimensional function optimization problems. The evolutionary algorithm that is selected in this paper is the Fast Evolutionary Programming (FEP) which uses Cauchy mutation to improve over the basic Gaussian mutation scheme. The swarm intelligence algorithm that is selected is the Artificial Bee Colony (ABC) algorithm which has been introduced recently and found to be very effective on many continuous optimization problems. This paper compares the performance of these two algorithms on a common set of benchmark problems in order to achieve a better understanding of their algorithmic nature and characteristics. The experimental results show that the performance of ABC is usually better than FEP, especially on complex multimodal functions, because ABC can deal with the problems of premature convergence and fitness stagnation more effectively than FEP.