Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Federated Linear Dueling Bandits
Xuhan Huang, Yan Hu, Zhiyan Li, Zhiyong Wang, Zhongxiang Dai
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 |
|---|---|---|---|
| add federated collaboration via FLDB-OGD over isolated per-agent LDB | linear dueling bandit · Synthetic and MovieLens | average cumulative regret | Improved |
| use exact federated GD (FLDB-GD) instead of one online GD step (FLDB-OGD) | FLDB-OGD · Synthetic (K=10, d=5) | average cumulative regret | Improved |
| use FLDB-OGD instead of FLDB-GD for runtime | FLDB-GD · Synthetic (K=10, d=5, N=100) | runtime per trial: > 9 h but < 10 h -> 70.25 s | Improved |
| increase number of agents N (ablation) | FLDB-OGD · Synthetic (K=10,50; d=5) | average cumulative regret | Improved |
| increase reward heterogeneity sigma from 0.1 to 0.5 (ablation) | FLDB-OGD and FLDB-GD · Synthetic heterogeneous (K=5, d=5) | average cumulative regret | Hurt |
| increase local update period tau (fewer communication rounds) (ablation) | FLDB-OGD · Synthetic (N=100, K=10, d=5, T=500) | final cumulative regret | Hurt |
| use tau=2 instead of tau=1 (ablation) | FLDB-OGD · Synthetic (N=100, K=10, d=5) | runtime per trial: 70.25 s -> 68.34 s | No clear effect |
Closest papers by meaning
- Communication-Efficient Federated Non-Linear Bandit Optimization (ICLR 2024 · similarity 0.89)
- Neural Dueling Bandits: Preference-Based Optimization with Human Feedback (ICLR 2025 · similarity 0.89)
- Online Clustering of Dueling Bandits (ICML 2025 · similarity 0.88)
- Variance-aware Regret Bounds for Stochastic Contextual Dueling Bandits (ICLR 2024 · similarity 0.88)
- Federated Combinatorial Multi-Agent Multi-Armed Bandits (ICML 2024 · similarity 0.88)
- Borda Regret Minimization for Generalized Linear Dueling Bandits (ICML 2024 · 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.
