Research map / Reinforcement learning / Policy gradient and actor-critic
Hybrid Learning with New Value Function for the Maximum Common Induced Subgraph Problem
Yanli Liu, Jiming Zhao, Chu-Min Li, Hua Jiang, Kun He
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 |
|---|---|---|---|
| Use DAL rewards that combine upper-bound reduction with the number of resulting domains instead of upper-bound reduction alone (ablation) | McSplit+RL and McSplit+LL · MCS ablation comparison instances | instances solved: McSplitRLD solves 138 more instances; 8.2%; McSplitLLD solves 154 more instances; 8.4% | Improved |
| Add hybrid RL/DAL branching to the DAL value function (ablation) | McSplitRLD and McSplitLLD · MCS ablation comparison instances | instances solved: McSplitDAL solves 138 (299) more instances than McSplitLLD (McSplitRLD) | Improved |
| Switch RL and DAL policies using the solution-size-related threshold rather than random selection (ablation) | McSplitDAL · 2,277 instances | instances solved: 121 more instances | Improved |
| Switch RL and DAL policies using the solution-size-related threshold rather than fixed tree depth (ablation) | McSplitDAL · 2,277 instances | instances solved: 286 more instances | Improved |
| Use DAL value function and hybrid RL/DAL branching instead of McSplit+LL branching | McSplit · Remaining 2,229 instances | instances solved: 292 more instances; 16.3% | Improved |
| Use DAL value function and hybrid RL/DAL branching instead of McSplit+RL branching | McSplit · Remaining 2,229 instances | instances solved: 437 more instances | Improved |
Closest papers by meaning
- A Learning Based Branch and Bound for Maximum Common Subgraph Related Problems (AAAI 2020 · similarity 0.93)
- Neural Graduated Assignment for Maximum Common Edge Subgraphs (ICLR 2026 · similarity 0.86)
- Planning in Branch-and-Bound: Model-Based Reinforcement Learning for Exact Combinatorial Optimization (AAAI 2026 · similarity 0.86)
- Exploratory Combinatorial Optimization with Reinforcement Learning (AAAI 2020 · similarity 0.85)
- SORREL: Suboptimal-Demonstration-Guided Reinforcement Learning for Learning to Branch (AAAI 2025 · similarity 0.84)
- Bayesian Optimization of Functions over Node Subsets in Graphs (NeurIPS 2024 · similarity 0.82)
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.
