Research map / Reinforcement learning / Policy gradient and actor-critic
Adaptive Estimator Selection for Off-Policy Evaluation
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 |
|---|---|---|---|
| Select bandwidth from the candidate set with Slope instead of using any one fixed bandwidth | Continuous contextual bandit kernel estimator · Continuous contextual bandits, 480 conditions | MSE | Helped |
| Select the partial importance-weighting false horizon with Slope | RL off-policy evaluation · RL, 106 experimental conditions | normalized MSE | Helped |
| Select the partial importance-weighting false horizon with Slope rather than Magic | RL off-policy evaluation · RL, 106 experimental conditions | statistically significant pairwise MSE comparison: 16% of the conditions; 25% | Mixed |
| Select the false horizon with Slope rather than using direct modeling alone | RL off-policy evaluation · Hybrid domain | MSE | Helped |
| Select the false horizon with Slope rather than using full-trajectory self-normalized doubly robust estimation | RL off-policy evaluation · Hybrid domain | MSE | Helped |
| Select the false horizon with Slope rather than using Magic | RL off-policy evaluation · Hybrid domain | MSE | Helped |
Closest papers by meaning
- Optimal and Adaptive Off-policy Evaluation in Contextual Bandits (ICML 2017 · similarity 0.92)
- More Efficient Off-Policy Evaluation through Regularized Targeted Learning (ICML 2019 · similarity 0.92)
- Efficient Policy Evaluation with Offline Data Informed Behavior Policy Design (ICML 2024 · similarity 0.89)
- Abstract Reward Processes: Leveraging State Abstraction for Consistent Off-Policy Evaluation (NeurIPS 2024 · similarity 0.89)
- Log-Sum-Exponential Estimator for Off-Policy Evaluation and Learning (ICML 2025 · similarity 0.89)
- Logarithmic Smoothing for Pessimistic Off-Policy Evaluation, Selection and Learning (NeurIPS 2024 · similarity 0.89)
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.