Research map / Bandits and online learning / Multi-armed bandits
Best Arm Identification in Linear Bandits with Linear Dimension Dependency
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 Y-ELIMTIL1 instead of XY-Allocation | Linear bandit best arm identification · Synthetic Data Set 1, d=10 | runtime (s): 1869.0 -> 0.2 | Helped |
| Use ALBA instead of XY-Adaptive | Linear bandit best arm identification · Synthetic Data Set 1, d=10 | runtime (s): 44.4 -> 0.3 | Helped |
| Increase dimension d from 2 to 10 for XY-Allocation | XY-Allocation · Synthetic Data Set 1 | runtime (s): 11.3 -> 1869.0 | Hurt |
| Increase dimension d from 2 to 10 for Y-ELIMTIL1 | Y-ELIMTIL1 · Synthetic Data Set 1 | runtime (s): 0.2 -> 0.2 | No clear effect |
| Use Y-ELIMTIL1/ALBA where baselines cannot finish | Y-ELIMTIL1 and ALBA · Synthetic Data Set 2 (d=60-100) | termination | Helped |
| Use Y-ELIMTIL1/ALBA where baselines cannot finish | Y-ELIMTIL1 and ALBA · Real-world Data Set (d=60-100) | termination | Helped |
| Use ALBA instead of Y-ELIMTIL1 on random vectors, d=70 | ALBA · Synthetic Data Set 2 | #samples used / (d*Delta2^-2): 318.94 -> 259.73 | Helped |
| Use ALBA instead of Y-ELIMTIL1 on random vectors, d=100 | ALBA · Synthetic Data Set 2 | #samples used / (d*Delta2^-2): 383.91 -> 302.99 | Helped |
| Use ALBA instead of Y-ELIMTIL1 on advertisement data, d=100 | ALBA · Real-world Data Set | #samples used / (d*Delta2^-2) (x10^-3): 0.84 -> 0.32 | Hurt |
| Use ALBA vs Y-ELIMTIL1 in runtime on real-world data | ALBA · Real-world Data Set | runtime (s): 18.6 -> 22.7 | Hurt |
| Use ALBA vs Y-ELIMTIL1 in runtime on random vectors, d=60 | ALBA · Synthetic Data Set 2 | runtime (s): 5.2 -> 5.6 | Hurt |
| Lower reward gap Delta2 (d=60 real data, 3.386) reduces sample efficiency | both algorithms · Real-world Data Set | number of samples | Hurt |
Closest papers by meaning
- Almost Minimax Optimal Best Arm Identification in Piecewise Stationary Linear Bandits (NeurIPS 2024 · similarity 0.92)
- Best Arm Identification in Generalized Linear Bandits via Hybrid Feedback (NeurIPS 2026 · similarity 0.92)
- Breaking the $\log(1/\Delta_2)$ Barrier: Better Batched Best Arm Identification with Adaptive Grids (ICLR 2025 · similarity 0.92)
- Optimal Multi-Fidelity Best-Arm Identification (NeurIPS 2024 · similarity 0.92)
- Balancing Performance and Costs in Best Arm Identification (NeurIPS 2025 · similarity 0.91)
- Near Optimal Best Arm Identification for Clustered Bandits (ICML 2025 · similarity 0.90)
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.