Generalize a Small Pre-trained Model to Arbitrarily Large TSP Instances
Shenzhen Institute of Artificial Intelligence and Robotics for Society(深圳市人工智能与机器人研究院);Institute of Robotics and Intelligent Manufacturing, The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳)机器人与智能制造研究院);School of Data Science, The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳)数据科学学院)
For the traveling salesman problem (TSP), the existing supervised learning based algorithms suffer seriously from the lack of generalization ability. To overcome this drawback, this paper tries to train (in supervised manner) a small-scale model, which could be repetitively used to build heat maps for TSP instances of arbitrarily large size, based on a series of techniques such as graph sampling, graph converting and heat maps merging. Furthermore, the heat maps are fed into a reinforcement learning approach (Monte Carlo tree search), to guide the search of high-quality solutions. Experimental results based on a large number of instances (with up to 10,000 vertices) show that, this new approach clearly outperforms the existing machine learning based TSP algorithms, and significantly improves the generalization ability of the trained model.
Mechanistic Interpretability for Neural TSP Solvers
University of Washington, Michael G. Foster School of Business(华盛顿大学,迈克尔·G·福斯特商学院);University of Washington, Paul G. Allen School of Computer Science & Engineering(华盛顿大学,保罗·G·阿伦计算机科学与工程学院)
Solve routing problems with a residual edge-graph attention neural network
Southwest Jiaotong University(西南交通大学);Technology and Equipment of Rail Transit Operation and Maintenance Key Laboratory of Sichuan Province(四川省轨道交通运维技术与装备重点实验室);Auburn University at Montgomery(奥本大学蒙哥马利分校)
Learning to Solve Multiple-TSP with Time Window and Rejections via Deep Reinforcement Learning
Nanyang Technological University(南洋理工大学);Singapore Institute of Manufacturing Technology (SIMTech)(新加坡制造技术研究院(SIMTech));Agency for Science Technology and Research (A*STAR)(科学、技术与研究局(A*STAR));Shandong University(山东大学);Technische Universiteit Delft(代尔夫特理工大学)
University of Latvia(拉脱维亚大学);Institute of Mathematics and Computer Science at University of Latvia(拉脱维亚大学数学与计算机科学研究所);Institute of Electronics and Computer Science(电子与计算机科学研究所)
Graph Neural Network-based Algorithm Selection for the Traveling Salesman Problem: A Systematic Study of Cost and Rank Losses under Distinct Budget Regimes
A Game-Theoretic Approach for Improving Generalization Ability of TSP Solvers
University of Chinese Academy of Sciences(中国科学院大学);King’s College London(伦敦国王学院);University College London(伦敦大学学院);Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)
在本文中,我们在可训练的\emph{Solver}和\emph{Data Generator}之间引入了一个两人零和框架,以提高基于深度学习的TSP求解器的泛化能力。基于\textsl{Policy Space Response Oracle}(PSRO)方法,我们的两人框架输出了一个最佳响应的求解器群体,我们可以混合并输出一个组合模型,该模型实现了对生成器的最小可利用性,从而在不同的TSP任务上实现了最具普遍性的性能。我们对不同类型和大小的TSP实例进行了实验。结果表明,我们的求解器即使在求解器从未遇到的任务上也能实现最先进的性能,而其他基于深度学习的求解器的性能由于过度拟合而急剧下降。为了证明我们的框架的原理,我们研究了所提出的两人游戏的学习结果,并证明了在训练过程中求解器种群的可利用性降低,并且它最终与生成器一起接近纳什均衡。
Learning-Based TSP-Solvers Tend to Be Overly Greedy
Academy of Mathematics and Systems Science, Chinese Academy of Sciences(中国科学院数学与系统科学研究院);School of Mathematics Sciences, University of Chinese Academy of Sciences(中国科学院大学数学科学学院)
Quantum Approaches to Urban Logistics: From Core QAOA to Clustered Scalability
Politecnico di Milano(米兰理工大学);International Foundation Big Data And Artificial Intelligence For Human Development Via Galliera n. 32 – Bologna(大数据与人工智能促进人类发展的国际基金会,via Galliera n. 32 – 玻利维亚)
NeuFACO: Neural Focused Ant Colony Optimization for Traveling Salesman Problem
College of Engineering and Computer Science, VinUniversity(工程与计算机科学学院,Vin大学);School of Information and Communications Technology, Hanoi University of Science and Technology(信息与通信技术学院,河内科学技术大学);University of Engineering and Technology, Vietnam National University(工程大学,越南国家大学)