大型候选解集子集选择基准测试套件

本文介绍了一个专为进化多目标优化领域设计的大型候选解集子集选择基准测试套件。该数据集由288个不同的候选解集组成，这些解集是通过直接在Pareto前沿上采样或通过运行多目标优化算法在特定测试问题上生成的。数据集旨在为研究人员提供一个平台，以理解和比较不同的子集选择方法，并开发新的优化算法。该数据集的应用领域包括但不限于环境选择、特征选择和传感器布置等，旨在通过高效的子集选择方法提高多目标优化算法的性能。

This paper presents a large-scale benchmark suite for subset selection of candidate solution sets, specifically tailored for the field of evolutionary multi-objective optimization. This dataset comprises 288 distinct candidate solution sets, which are generated either via direct sampling on the Pareto front or by executing multi-objective optimization algorithms on specific test problems. The dataset aims to provide researchers with a platform to understand and compare diverse subset selection methods, as well as develop novel optimization algorithms. Application areas of this dataset include, but are not limited to, environmental selection, feature selection, and sensor placement, among others, with the core objective of enhancing the performance of multi-objective optimization algorithms through efficient subset selection approaches.