BriefGPT.xyz
Nov, 2024
学习混合整数线性规划的最优目标值
Learning optimal objective values for MILP
HTML
PDF
Lara Scavuzzo, Karen Aardal, Neil Yorke-Smith
TL;DR
本研究针对现代混合整数线性规划（MILP）求解器的一个空白，通过引入图神经网络（GNN）架构的预测方法，预测最优目标值或当前解是否最优。实验结果表明，该方法在预测任务中表现出高准确率，并超越了现有方法，为将机器学习驱动的预测集成到MILP求解器中提供了新的可能性。
Abstract
Modern Mixed Integer Linear Programming (
MILP
) solvers use the Branch-and-Bound algorithm together with a plethora of auxiliary components that speed up the search. In recent years, there has been an explosive development in the use of
→