首页> 外文会议>International symposium on experimental algorithms >Submodular Minimization in the Context of Modern LP and MILP Methods and Solvers
【24h】

Submodular Minimization in the Context of Modern LP and MILP Methods and Solvers

机译:现代LP和MILP方法和求解器背景下的亚模最小化

获取原文

摘要

We consider the application of mixed-integer linear programming (MILP) solvers to the minimization of submodular functions. We evaluate common large-scale linear-programming (LP) techniques (e.g., column generation, row generation, dual stabilization) for solving a LP reformulation of the submodular minimization (SM) problem. We present heuristics based on the LP framework and a MILP solver. We evaluated the performance of our methods on a test bed of min-cut and matroid-intersection problems formulated as SM problems.
机译:我们考虑将混合整数线性规划(MILP)求解器应用到最小化子模块函数中。我们评估了常见的大规模线性编程(LP)技术(例如列生成,行生成,双重稳定),以解决子模最小化(SM)问题的LP重新编制。我们介绍基于LP框架和MILP求解器的启发式方法。我们在制定为SM问题的最小切割和拟阵相交问题的测试台上评估了我们方法的性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号