Proceedings of the
The Nineteenth International Conference on Computational Intelligence and Security (CIS 2023)
December 1 – 4, 2023, Haikou, China
An Effective Indicator to Evaluate the Congestion of Individual: Compound Congestion Indicator
School of Applied Mathematics, Guangdong University of Technology, China.
ABSTRACT
In low-dimensional cases, individuals selected by NSGA-II have poor diversity. To this end, instead of NSGAII's crowding distance, we combine the spanning angle and radius contribution and then construct a compound congestion indicator to evaluate an individual's quality. Studies show that this indicator can more effectively elect excellent individuals. These individuals converge faster and disseminate more evenly. Additionally, the individual at the boundary of the Pareto front performs better than others. However, the local greed strategy of NSGA-II can't ensure to selection of these individuals. Therefore, we use the global greed strategy to achieve that. We embed the proposed indicator and the global greed strategy into NSGA-II. Numerical experiments show that the performance of the proposed algorithm is better than that of NSGA-II.
Keywords: Compound congestion indicator, NSGA-II, Multi-objective optimization, Evolutionary algorithm.
Download PDF
School of Applied Mathematics, Guangdong University of Technology, China.
ABSTRACT
In low-dimensional cases, individuals selected by NSGA-II have poor diversity. To this end, instead of NSGAII's crowding distance, we combine the spanning angle and radius contribution and then construct a compound congestion indicator to evaluate an individual's quality. Studies show that this indicator can more effectively elect excellent individuals. These individuals converge faster and disseminate more evenly. Additionally, the individual at the boundary of the Pareto front performs better than others. However, the local greed strategy of NSGA-II can't ensure to selection of these individuals. Therefore, we use the global greed strategy to achieve that. We embed the proposed indicator and the global greed strategy into NSGA-II. Numerical experiments show that the performance of the proposed algorithm is better than that of NSGA-II.
Keywords: Compound congestion indicator, NSGA-II, Multi-objective optimization, Evolutionary algorithm.
Download PDF