awesome-ml4co

Why we included this project

Anyone trying to apply machine learning to combinatorial optimization problems will find this list hard to beat as a starting point. It collects papers that use learning-based methods on classic problems like the traveling salesman problem, vehicle routing, job-shop scheduling, graph matching, and SAT, and organizes them by problem type rather than by technique. That organization matters in practice: if you are facing a specific optimization task, you can jump straight to the relevant literature instead of digging through unrelated work. The list is maintained by the SJTU Thinklab research group and takes community contributions, so it stays reasonably current and includes both surveys and recent method papers. For someone scoping a research project or deciding whether learned solvers are worth trying on a particular problem, it saves a lot of scattered searching.