Optimal Distribution of Solutions for Crowding Distance on Linear Pareto Fronts of Two-Objective Optimization Problems

2025-04-24

AI Summary

This work addresses the problem of optimal solution distribution on a bi-objective linear Pareto front under the crowding distance metric, aiming to maximize the minimum crowding distance and filling a theoretical gap in NSGA-II analysis. Methodologically, we conduct rigorous theoretical analysis combined with computational experiments, proving for the first time that uniform distribution is not optimal with respect to crowding distance. We identify and characterize NSGA-II’s implicit tendency to duplicate extreme solutions at the front boundaries—a phenomenon detrimental to optimality. To address this, we propose an enhanced NSGA-II employing a (μ+1) generational update mechanism, and theoretically derive the structural properties of the optimal distribution on linear fronts. Experimental results confirm that the algorithm converges to a high-quality, near-uniform distribution—exhibiting only boundary duplications—thereby substantially improving both distribution quality and theoretical interpretability.

Technology Category

• Search and Optimization: Mixed Discrete/Continuous Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Game Theory and Economic Paradigms: Fair Division

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Characteristics of an evolutionary multi-objective optimization (EMO) algorithm can be explained using its best solution set. For example, the best solution set for SMS-EMOA is the same as the optimal distribution of solutions for hypervolume maximization. For NSGA-III, if the Pareto front has intersection points with all reference lines, all of those intersection points are the best solution set. For MOEA/D, the best solution set is the set of the optimal solution of each sub-problem. Whereas these EMO algorithms can be analyzed in this manner, the best solution set for the most well-known and frequently-used EMO algorithm NSGA-II has not been discussed in the literature. This is because NSGA-II is not based on any clear criterion to be optimized (e.g., hypervolume maximization, distance minimization to the nearest reference line). As the first step toward the best solution set analysis for NSGA-II, we discuss the optimal distribution of solutions for the crowding distance under the simplest setting: the maximization of the minimum crowding distance on linear Pareto fronts of two-objective optimization problems. That is, we discuss the optimal distribution of solutions on a straight line. Our theoretical analysis shows that the uniformly distributed solutions are not the best solution set. However, it is also shown by computational experiments that the uniformly distributed solutions (except for the duplicated two extreme solutions at each edge of the Pareto front) are obtained from a modified NSGA-II with the ($mu$ + 1) generation update scheme.

Problem

• Analyzes NSGA-II's best solution set for crowding distance
• Examines optimal solution distribution on linear Pareto fronts
• Compares uniform distribution with theoretical optimal solutions

Innovation

• Analyzes NSGA-II's best solution set distribution
• Maximizes minimum crowding distance on linear fronts
• Uses modified NSGA-II with (μ + 1) scheme