Self-generated Koopmans-Beckmann QAP Instances
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/PKUTAN/SAWT
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了自生成的Koopmans-Beckmann QAP实例，通过随机抽取位置节点坐标和流量指标来创建。坐标均匀地从单位正方形[0, 1]^2内抽取，而流量指标则是随机均匀采样，在特定条件下具有概率p=0.7。该数据集的规模可达到5120个训练实例和256个测试实例，其任务是进行二次分配问题（QAP）优化。
This dataset contains self-generated Koopmans-Beckmann Quadratic Assignment Problem (QAP) instances, which are constructed by randomly sampling the coordinates of location nodes and corresponding flow metrics. The coordinates are uniformly sampled from the unit square [0, 1]^2, while the flow metrics are randomly and uniformly sampled with a probability of p=0.7 under specific conditions. The dataset has a maximum size of 5120 training instances and 256 test instances, with its core task being Quadratic Assignment Problem (QAP) optimization.
提供机构：
Authors of the paper
