Using an adaptation of a binary search tree to improve the NSGA-II nondominated sorting procedure

Abstract

In this paper, we propose an adaptation to Nondominated Sorting Genetic Algorithm (NSGA-II), introducing a data structure, called NonDominated Tree (NDT). The NDT is an adaptation of a Binary Search Tree and is used to identify the nondominated fronts in only one run. This structure may be used to improve even more the performance of NSGA-II and other Evolutionary Algorithms (EAs) that use nondominated sorting procedures. It reduces the number of comparisons performed by the NSGA-II nondominated sorting routine. Some tests demonstrated that the proposed structure improves the search of fronts of nondominated solutions in an efficient way. © 2010 Springer-Verlag.