An efficient nondominated sorting algorithm
Junchen Wang, Changhe Li, Yiya Diao, Sanyou Y. Zeng, Hui Wang · Proceedings of the Genetic and Evolutionary Computation Conference Companion · 2018
Nondominated sorting (NS) is commonly needed in multi-objective optimization to distinguish the fitness of solutions. Since it was suggested, several NS algorithms have been proposed to reduce its time complexity. In our study, we found that their performances are closely related to properties of the distribution of a data set, especially the number of fronts. To address this issue, we propose a novel NS algorithm Filter Sort. We also propose a new benchmark data generator for evaluating the performance of a NS algorithm. Experimental results show that our algorithm is superior to several state-of-the-art NS algorithms in most cases.