Research map / Multi-agent systems and games / Cooperative multi-agent RL
Adaptive and Robust DBSCAN With Multi-Agent Reinforcement Learning
Hao Peng, Xiang Huang, Shuo Sun, Ruitong Zhang, Xizhao Wang, Philip S. Yu
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 |
|---|---|---|---|
| allocate density partitions to separate agents (AR-DBSCAN vs DRL-DBSCAN) | DRL-DBSCAN · 9 offline datasets (average) | NMI: .820 -> .891 | Improved |
| allocate density partitions to separate agents (AR-DBSCAN vs DRL-DBSCAN) | DRL-DBSCAN · 9 offline datasets (average) | ARI: .749 -> .826 | Improved |
| allocate density partitions to separate agents | DRL-DBSCAN · Pathbased and Cure-t2 | ARI: underperforms DRL-DBSCAN with 0.041 and 0.071 on the ARI metric | Hurt |
| add edge weight normalization (ablation) | AR-DBSCAN · Pathbased | NMI: 11.9% improvement; .737 -> .825 | Improved |
| add one-dimensional structural entropy normalization (ablation) | AR-DBSCAN · Pathbased | NMI: 5.4% improvement; .783 -> .825 | Improved |
| add edge weight normalization (ablation) | AR-DBSCAN · Cure-t2 | NMI: .938 -> .764 | Hurt |
| add edge weight normalization (ablation) | AR-DBSCAN · D31 | NMI: .729 -> .773 | Improved |
| add H1(G) normalization (ablation) | AR-DBSCAN · Cure-t2 | NMI: .757 -> .764 | Improved |
| add DRL parameter search and agents on top of encoding-tree clustering (ablation) | Optimal two-level encoding tree · Pathbased | NMI: .913 -> .825 | Hurt |
| add DRL parameter search and agents on top of encoding-tree clustering (ablation) | Optimal two-level encoding tree · Aggregation | NMI: .836 -> .978 | Improved |
| add DRL parameter search and agents on top of encoding-tree clustering (ablation) | Optimal two-level encoding tree · Cure-t2 | NMI: .561 -> .764 | Improved |
| use AR-DBSCAN instead of other search baselines | baseline parameter search · offline datasets | convergence rounds: converges within 15 rounds and about two times speedup | Improved |
Closest papers by meaning
- Multi-View Clustering by Inter-cluster Connectivity Guided Reward (ICML 2024 · similarity 0.85)
- HYGMA: Hypergraph Coordination Networks with Dynamic Grouping for Multi-Agent Reinforcement Learning (ICML 2025 · similarity 0.84)
- MAT-Agent: Adaptive Multi-Agent Training Optimization (NeurIPS 2025 · similarity 0.84)
- Just Cluster It: An Approach for Exploration in High-Dimensions using Clustering and Pre-Trained Representations (ICML 2024 · similarity 0.84)
- Meta-Reinforcement Learning via Exploratory Task Clustering (AAAI 2024 · similarity 0.83)
- Online Clustering of Dueling Bandits (ICML 2025 · similarity 0.83)
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.
