Research map / Reinforcement learning / Visual and representation RL
A Deep Reinforcement Learning Framework for Column Generation
Cheng Chi, Amine Mohamed Aboussalah, Elias Boutros Khalil, Juyoung Wang, Zoha Sherkat-Masoumi
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 greedy column selection with RLCG | column generation · CSP n=750 | mean CG iterations: 292.8 -> 227.2 | Improved |
| replace greedy column selection with RLCG | column generation · CSP n=50 | mean CG iterations: 51.2 -> 41.6 | Improved |
| replace greedy column selection with RLCG | column generation · VRPTW large | mean CG iterations: 128.1 -> 75.6 | Improved |
| replace greedy column selection with RLCG | column generation · VRPTW small | mean CG iterations: 18.8 -> 9.5 | Improved |
| replace MILP one-step-lookahead expert with RLCG | column generation · CSP n=750 | mean time (s): 827.9 -> 460.3 | Improved |
| replace MILP expert with RLCG | column generation · CSP n=50 | mean CG iterations: 42.5 -> 41.6 | Mixed |
| train with curriculum (easy to hard) rather than random order (ablation) | RLCG · CSP training | CG iterations during training | Improved |
| weight objective decrease in reward (alpha=300 vs alpha=0) (ablation) | RLCG · CSP validation | greedy/RL iteration ratio: 0.91 -> 1.25 | Improved |
| different random seeds for best model (ablation) | RLCG · CSP validation | greedy/RL iteration ratio: 1.23 to 1.25 | No clear effect |
Closest papers by meaning
- FFCG: Effective and Fast Family Column Generation for Solving Large-Scale Linear Program (AAAI 2025 · similarity 0.93)
- A Reinforcement-Learning-Based Multiple-Column Selection Strategy for Column Generation (AAAI 2024 · similarity 0.93)
- Reinforcement Learning for Integer Programming: Learning to Cut (ICML 2020 · similarity 0.89)
- Adaptive Stabilization Based on Machine Learning for Column Generation (ICML 2024 · similarity 0.87)
- Learn2Aggregate: Supervised Generation of Chvatal-Gomory Cuts Using Graph Neural Networks (AAAI 2025 · similarity 0.86)
- Gradient Boosting Reinforcement Learning (ICML 2025 · similarity 0.85)
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.
