Research map / Bandits and online learning / Contextual bandits and Thompson sampling
Bi-Level Contextual Bandits for Individualized Resource Allocation Under Delayed Feedback
Mohammadsina Almasi, Hadis Anahideh
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 bi-level MetaCUB instead of UCB for allocation | MetaCUB · ELS Black, delayed | subgroup allocation ratio: 0.33 -> 1.00 | Improved |
| use MetaCUB instead of EXP3 | MetaCUB · ELS Asian, delayed | subgroup allocation ratio: 0.27 -> 1.02 | Improved |
| use MetaCUB instead of UCB | MetaCUB · ELS White, delayed | subgroup allocation ratio: 1.42 -> 0.96 | Improved |
| use MetaCUB instead of UCB | MetaCUB · JOBS Black, delayed | subgroup allocation ratio: 0.54 -> 0.98 | Improved |
| use MetaCUB instead of DUCB | MetaCUB · JOBS Black, delayed | subgroup allocation ratio: 1.29 -> 0.98 | Improved |
| use MetaCUB instead of SWUCB | MetaCUB · JOBS Black, delayed | subgroup allocation ratio: 1.30 -> 0.98 | Improved |
| use MetaCUB instead of EXP3 | MetaCUB · JOBS Black, immediate | subgroup allocation ratio: 1.20 -> 1.03 | Improved |
| use MetaCUB instead of UCB | MetaCUB · ELS Asian, immediate | subgroup allocation ratio: 0.62 -> 0.84 | Improved |
| use MetaCUB instead of baselines | MetaCUB · ELS and JOBS cumulative regret | cumulative regret | Improved |
| use nonlinear (neural) reward model instead of linear | all algorithms · ELS and JOBS cumulative regret | cumulative regret | Improved |
| use Type-I (peaked) instead of Type-II (flat) delay kernels | all algorithms · ELS and JOBS cumulative regret | cumulative regret | Improved |
| add delay-aware kernel modeling and subgroup structure | MetaCUB · delayed feedback regret | cumulative regret | Improved |
Closest papers by meaning
- The Hidden Cost of Waiting for Accurate Predictions (ICLR 2025 · similarity 0.86)
- MultiScale Contextual Bandits for Long Term Objectives (NeurIPS 2025 · similarity 0.85)
- Allocation Requires Prediction Only if Inequality Is Low (ICML 2024 · similarity 0.85)
- Context in Public Health for Underserved Communities: A Bayesian Approach to Online Restless Bandits (AAAI 2025 · similarity 0.84)
- Cost-Aware Near-Optimal Policy Learning (AAAI 2025 · similarity 0.84)
- A Continuous-time Tractable Model for Present-biased Agents (AAAI 2025 · similarity 0.84)
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.
