ILSGVCP: An improved local search algorithm for generalized vertex cover problem
针对广义顶点覆盖问题，提出一种改进的局部搜索算法ILSGVCP，通过新的初始化规则和动态扰动机制提升求解质量，在随机实例和DIMACS实例上优于现有算法。
The generalized vertex cover problem (GVCP) is a well-known combinatorial problem of the classic minimum vertex cover problem with rich applications. And local search algorithm is an effective heuristic algorithm to solve GVCP. Therefore, In this paper, we develop an improved local search algorithm for the problem, namely ILSGVCP. Specifically, new choosing small weight rules called CSWR is presented for initialization to acquire a high-quality candidate solution. Furthermore, we propose a dynamic perturbation mechanism called DP with a novel formula which makes local search perturbed dynamically. Moreover, we combine CSWR and DP to propose our algorithm ILSGVCP. Finally, we carry out experiments to evaluate ILSGVCP on random instances with up to 1000 vertices and 400,000 edges and DIMACS instances. A detailed experimental evaluation reveals that ILSGVCP outperforms other state-of-the-art algorithms for the GVCP.