Column Generation

Two-Stage Learning to Branch in Branch-Price-and-Cut Algorithms for Solving Vehicle Routing Problems Exactly

Branching is one of the most important components in branch-price-and-cut (BPC) algorithms for solving vehicle routing problems (VRPs) exactly. However, learning to branch is much more challenging in BPC than in branch-and-cut algorithms that are used for solving general mixed integer programs. To address such challenges, we propose the first effective learning-to-branch framework in BPC algorithms, leading to a novel two-stage learning-based branching (2LBB) strategy.

L-DDOIs: Learning-Based Deep Dual-Optimal Inequalities for Stabilizing Column Generation

Stay tuned.

L-DDOIs: Learning-Based Deep Dual-Optimal Inequalities for Stabilizing Column Generation

Conference presentations on learning-based stabilization for column generation at the 2025 INFORMS Annual Meeting and the 2026 Transportation Science & Logistics Conference.

Accelerating Dynamic Programming via Dual Selection for Column-Generation-Based Exact Solution Frameworks

Presented at the 2024 INFORMS Optimization Society Conference with travel support.

Learning to Branch with Column Generation

Presented at the 2022 INFORMS Annual Meeting on early learning-to-branch ideas in column-generation settings.