Reducing the Computational Cost in Multi-objective Evolutionary Algorithms by Filtering Worthless Individuals
The large number of exact fitness function evaluations makes evolutionary algorithms to have computational cost (especially in Multi Objective Problems (MOPs)). In some real- world problems, reducing number of these evaluations is much more valuable even by increasing computational complexity and spending more time. To fulfil this target, we introduce an effective factor, in spite of applied factor in Adaptive Fuzzy Fitness Granulation NSGA#1030;#1030; (AFFG_NSGA#1030;#1030;) algorithm, to filter out worthless individuals more precisely. Our proposed approach is compared with respect to AFFG_NSGA#1030;#1030;, using the Hypervolume (HV) and the Inverted Generational Distance (IGD) performance measures. The proposed method is applied to 1 traditional and 1 state-of-the-art benchmarks with considering 3 different dimensions. From an average performance view, the results indicate that although decreasing the number of fitness evaluations leads to have performance reduction but it is not tangible compared to what we gain.
Keywords: Multi objective evolutionary algorithm optimization, Fitness approximation, Information granulation, Pareto optimal solutions.
Download Full-Text
ABOUT THE AUTHORS
Zahra Pourbahman
Z. Pourbahman is the MSc student in artificial intelligence of CSE and IT Department of Shiraz University. As one of her research interests, she focuses on evolutionary algorithms specially fitness approximation domain. Also, she is co-author of several articles in Machine Translation area especially in Natural Language Processing domain. She was programmer team leader in Telecommunication Company of Iran in 2008.
Ali Hamzeh
A. Hamzeh received his Ph.D. in artificial intelligence from Iran University of Science and Technology (IUST) in 2007. Since then, he has been working as assistant professor in CSE and IT Department of Shiraz University. There, he is one of the founders of local CERT center which serves as the security and protection service provider in its local area. As one of his research interests, he recently focuses on cryptography and steganography area and works as a team leader in CERT center to develop and break steganography method, especially in image spatial domain. Also, he works as one of team leaders of Soft Computing group of shiraz university working on bio-inspired optimization algorithms. He is co-author of several articles in security and optimization.
Zahra Pourbahman
Z. Pourbahman is the MSc student in artificial intelligence of CSE and IT Department of Shiraz University. As one of her research interests, she focuses on evolutionary algorithms specially fitness approximation domain. Also, she is co-author of several articles in Machine Translation area especially in Natural Language Processing domain. She was programmer team leader in Telecommunication Company of Iran in 2008.
Ali Hamzeh
A. Hamzeh received his Ph.D. in artificial intelligence from Iran University of Science and Technology (IUST) in 2007. Since then, he has been working as assistant professor in CSE and IT Department of Shiraz University. There, he is one of the founders of local CERT center which serves as the security and protection service provider in its local area. As one of his research interests, he recently focuses on cryptography and steganography area and works as a team leader in CERT center to develop and break steganography method, especially in image spatial domain. Also, he works as one of team leaders of Soft Computing group of shiraz university working on bio-inspired optimization algorithms. He is co-author of several articles in security and optimization.
