Thinklab-SJTU/awesome-ml4co

Curated repository organizing 30+ combinatorial optimization problem categories—from classical (TSP, VRP, knapsack) to emerging domains (EDA, causal discovery, predict+optimize)—with categorized papers spanning neural networks, graph neural networks, and reinforcement learning approaches. Covers both survey papers and problem-specific solutions, emphasizing research from SJTU Thinklab alongside community contributions to bridge discrete optimization and modern deep learning paradigms.

2,087 stars.