Research map / Bandits and online learning / Multi-armed bandits
Ultra Fast Medoid Identification via Correlated Sequential Halving
Tavor Baharav, David Tse
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 correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · RNA-Seq 20k, ℓ1 | pulls per arm: 121 (2.1%) -> 2.43 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · RNA-Seq 100k, ℓ1 | pulls per arm: 420 -> 2.10 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · Netflix 20k, cosine distance | pulls per arm: 85.8 -> 15.0 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · Netflix 100k, cosine distance | pulls per arm: 90.5 (6%) -> 18.5 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · MNIST Zeros, ℓ2 | pulls per arm: 91.2 (.1%) -> 47.9 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · RNA-Seq 20k, ℓ1 | time: 246 -> 10.9 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · RNA-Seq 100k, ℓ1 | time: 5819 -> 64.2 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · Netflix 20k, cosine distance | time: 593 -> 6.82 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · Netflix 100k, cosine distance | time: 6495 -> 53.4 | Improved |
| Use correlated reference-point sampling with sequential halving instead of Med-dit's adaptive bandit sampling | Medoid identification · MNIST Zeros, ℓ2 | time: 151 -> 1.46 | Improved |
Closest papers by meaning
- OneBatchPAM: A Fast and Frugal K-Medoids Algorithm (AAAI 2025 · similarity 0.85)
- Near Optimal Best Arm Identification for Clustered Bandits (ICML 2025 · similarity 0.83)
- Optimal Multi-Fidelity Best-Arm Identification (NeurIPS 2024 · similarity 0.82)
- Almost Minimax Optimal Best Arm Identification in Piecewise Stationary Linear Bandits (NeurIPS 2024 · similarity 0.82)
- New Classes of the Greedy-Applicable Arm Feature Distributions in the Sparse Linear Bandit Problem (AAAI 2024 · similarity 0.82)
- On Universally Optimal Algorithms for A/B Testing (ICML 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.
