arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2312.04526cs.DM
收藏

全局支配问题的算法

Algorithms for the Global Domination Problem

Ernesto Parra Inza, Nodari Vakhania, Jose M. Sigarreta Almira, Frank A. Hernández Mira

AI总结:

针对此前缺乏精确与近似算法的全局支配问题,提出两种精确算法、三种启发式及净化过程,并通过大规模基准验证其效率与近优性。

摘要:

图 G 中的支配集 D 是其顶点的一个子集,使得图中不属于集合 D 的每个顶点都与集合 D 中的至少一个顶点相邻。如果图 G 的一个顶点集同时是图 G 及其补图的支配集,则该顶点集称为全局支配集。目标是找到基数最小的全局支配集。该问题已知为 NP 难问题,此前既不存在精确算法,也不存在近似算法。我们提出两种精确求解方法,其中一种基于整数线性规划(ILP)建模,三种启发式算法,以及一种特殊的净化过程,该过程可进一步减小由任一启发式算法给出的全局支配集的规模。我们证明该问题在受限图类型上仍为 NP 难,并指出若干图族,对这些图而言启发式算法可保证最优性。对于顶点数超过 230 且至多 1080 的图(这些是被最优求解的最大基准实例),第二种精确算法的速度约为 ILP 的两倍。启发式算法在现有 2284 个顶点数至多 14000 的基准问题实例上进行了测试,并在不到一分钟内为最大实例给出了解。值得注意的是,在已获得最优解的 1000 个实例中,约 52% 的实例至少有一种启发式算法生成了最优解,而其余实例的平均近似误差为 1.07%。

英文摘要:

A dominating set D in a graph G is a subset of its vertices such that every vertex of the graph which does not belong to set D is adjacent to at least one vertex from set D. A set of vertices of graph G is a global dominating set if it is a dominating set for both, graph G and its complement. The objective is to find a global dominating set with the minimum cardinality. The problem is known to be NP-hard. Neither exact nor approximation algorithm existed . We propose two exact solution methods, one of them being based on an integer linear program (ILP) formulation, three heuristic algorithms and a special purification procedure that further reduces the size of a global dominated set delivered by any of our heuristic algorithms. We show that the problem remains NP-hard for restricted types of graphs and specify some families of graphs for which the heuristics guarantee the optimality. The second exact algorithm turned out to be about twice faster than ILP for graphs with more than 230 vertices and up to 1080 vertices, which were the largest benchmark instances that were solved optimally. The heuristics were tested for the existing 2284 benchmark problem instances with up to 14000 vertices and delivered solutions for the largest instances in less than one minute. Remarkably, for about 52% of the 1000 instances with the obtained optimal solutions, at least one of the heuristics generated an optimal solution, where the average approximation error for the remaining instances was 1.07%.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑