Research map / Graph learning / Graph algorithms and clustering
What’s Wrong with Deep Learning in Tree Search for Combinatorial Optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace the GCN outputs with random values in the full tree-search configuration (ablation) | DGL-TreeSearch · SATLIB | MIS size: 426.37 (0.99) -> 426.48 (0.99) | No clear effect |
| Replace the GCN outputs with random values in the full tree-search configuration (ablation) | DGL-TreeSearch · DIMACS | MIS size: 76.57 (-) -> 76.51 (-) | No clear effect |
| Replace the GCN outputs with random values in the full tree-search configuration (ablation) | DGL-TreeSearch · ER 700-800 | MIS size: 43.91 (-) -> 44.02 (-) | No clear effect |
| Replace the GCN with random outputs (ablation) | Learning What To Defer · DIMACS | MIS size: 70.11 (-) -> 35.84 (-) | Hurt |
| Add graph reduction and local search (ablation) | DGL-TreeSearch · HRG 700-800 | MIS size: 221.80 (0.75) -> 304.21 (1.00) | Helped |
| Add weighted queue popping with queue pruning (ablation) | DGL-TreeSearch · HRG 700-800 | solutions found: 5 -> 98\% | Helped |
| Add weighted queue popping (ablation) | DGL-TreeSearch · VC-BM | time to solution: 139.55 -> 68.33 | Helped |
| Use the GCN rather than random outputs (ablation) | DGL-TreeSearch · REDDIT-B | MIS size: 186.31 -> 248.96 | Hurt |
| Use KaMIS rather than the tree-search solver | MIS solvers · SATLIB | time to solution: 7.99 -> 4.51 | Helped |
Closest papers by meaning
- Combinatorial Optimization with Graph Convolutional Networks and Guided Tree Search (NeurIPS 2018 · similarity 0.95)
- Learning to Explore and Exploit with GNNs for Unsupervised Combinatorial Optimization (ICLR 2025 · similarity 0.91)
- COMBHelper: A Neural Approach to Reduce Search Space for Graph Combinatorial Problems (AAAI 2024 · similarity 0.90)
- Towards General Algorithm Discovery for Combinatorial Optimization: Learning Symbolic Branching Policy from Bipartite Graph (ICML 2024 · similarity 0.90)
- Are Graph Neural Networks Optimal Approximation Algorithms? (NeurIPS 2024 · similarity 0.89)
- GOAL: A Generalist Combinatorial Optimization Agent Learner (ICLR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.