当前位置: X-MOL 学术Artif. Intell. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
SCCWalk: An efficient local search algorithm and its improvements for maximum weight clique problem
Artificial Intelligence ( IF 5.1 ) Pub Date : 2020-03-01 , DOI: 10.1016/j.artint.2019.103230
Yiyuan Wang , Shaowei Cai , Jiejiang Chen , Minghao Yin

Abstract The maximum weight clique problem (MWCP) is an important generalization of the maximum clique problem with wide applications. In this study, we develop two efficient local search algorithms for MWCP, namely SCCWalk and SCCWalk4L, where SCCWalk4L is improved from SCCWalk for large graphs. There are two main ideas in SCCWalk, including strong configuration checking (SCC) and walk perturbation. SCC is a new variant of a powerful strategy called configuration checking for local search. The walk perturbation procedure is used to lead the algorithm to leave the current area and come into a new area of feasible solution space. Moreover, to improve the performance on massive graphs, we apply a low-complexity heuristic called best from multiple selection to select the swapping vertex pair quickly and effectively, resulting in the SCCWalk4L algorithm. In addition, SCCWalk4L uses two recent reduction rules to decrease the scale of massive graphs. We carry out experiments to evaluate our algorithms on several popular benchmarks, which are divided into two groups, including classical benchmarks of small graphs namely DIMACS, BHOSLIB, winner determination problem, and graphs derived from clustering aggregation, as well as massive graphs, including a suite of massive real-world graphs and large-scale FRB graphs. Experiments show that, compared to state-of-the-art heuristic algorithms and exact algorithm, the proposed algorithms perform better on classical benchmarks, and obtain the best solutions for most massive graphs.

中文翻译:

SCCWalk:一种高效的局部搜索算法及其对最大权重团问题的改进

摘要 最大权团问题(MWCP)是最大团问题的重要推广,具有广泛的应用。在这项研究中,我们为 MWCP 开发了两种有效的局部搜索算法,即 SCCWalk 和 SCCWalk4L,其中 SCCWalk4L 是从 SCCWalk 改进而来的大图。SCCWalk 中有两个主要思想,包括强配置检查(SCC)和步行扰动。SCC 是一种称为本地搜索配置检查的强大策略的新变体。游走扰动过程用于引导算法离开当前区域,进入可行解空间的新区域。此外,为了提高海量图的性能,我们应用了一种称为多选最佳的低复杂度启发式方法来快速有效地选择交换顶点对,导致 SCCWalk4L 算法。此外,SCCWalk4L 使用了两个最近的缩减规则来减小海量图的规模。我们在几个流行的基准测试中进行了实验来评估我们的算法,这些基准分为两组,包括小图的经典基准,即 DIMACS、BHOSLIB、获胜者确定问题和从聚类聚合导出的图,以及大量图,包括大量真实世界图和大规模 FRB 图的套件。实验表明,与最先进的启发式算法和精确算法相比,所提出的算法在经典基准上的表现更好,并获得了大多数海量图的最佳解。我们在几个流行的基准测试中进行了实验来评估我们的算法,这些基准分为两组,包括小图的经典基准,即 DIMACS、BHOSLIB、获胜者确定问题和从聚类聚合导出的图,以及大量图,包括大量真实世界图和大规模 FRB 图的套件。实验表明,与最先进的启发式算法和精确算法相比,所提出的算法在经典基准上的表现更好,并获得了大多数海量图的最佳解。我们在几个流行的基准测试中进行了实验来评估我们的算法,这些基准分为两组,包括小图的经典基准,即 DIMACS、BHOSLIB、获胜者确定问题和从聚类聚合导出的图,以及大量图,包括大量真实世界图和大规模 FRB 图的套件。实验表明,与最先进的启发式算法和精确算法相比,所提出的算法在经典基准上的表现更好,并获得了大多数海量图的最佳解。包括一套大规模的真实世界图和大规模 FRB 图。实验表明,与最先进的启发式算法和精确算法相比,所提出的算法在经典基准上的表现更好,并获得了大多数海量图的最佳解。包括一套大规模的真实世界图和大规模 FRB 图。实验表明,与最先进的启发式算法和精确算法相比,所提出的算法在经典基准上的表现更好,并获得了大多数海量图的最佳解。
更新日期:2020-03-01
down
wechat
bug