Research map / Reinforcement learning / Policy gradient and actor-critic
CAB: Continuous Adaptive Blending for Policy Evaluation and Learning
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 CAB instead of DM | POEM and Propensity SVM-Rank · Letter | test error: 0.6372 -> 0.5740 | Helped |
| Use CAB instead of SB | POEM and Propensity SVM-Rank · Optdigits | test error: 0.0460 -> 0.0445 | Helped |
| Use CAB instead of cIPS | POEM and Propensity SVM-Rank · Satimage | test error: 0.2415 -> 0.2442 | Hurt |
| Use CAB instead of cIPS | POEM and Propensity SVM-Rank · Pendigits | test error: 0.1228 -> 0.0917 | Helped |
| Use CAB instead of SWITCH | Propensity SVM-Rank · Yahoo! LTR | average rank of relevant results: 10.95 -> 10.83 | Helped |
| Replace CAB's IPS component with clipped DR (CAB-DR) | POEM · Pendigits | test error: 0.0917 -> 0.0946 | Hurt |
| Increase the CAB clipping constant M | CAB evaluation · Satimage and Yahoo! LTR | bias | Helped |
| Increase the CAB clipping constant M | CAB evaluation · Satimage and Yahoo! LTR | variance | Hurt |
| Use CAB rather than cIPS, imputing a reward-model estimate instead of zero | Off-policy evaluation · UCI datasets and Yahoo! LTR | MSE | Helped |
| Use CAB rather than SB's static blending weight | Off-policy evaluation · Most tested datasets | MSE | Helped |
| Use CAB rather than DR | Off-policy evaluation · Most tested datasets | MSE | Helped |
| Vary the amount of data used to train the reward model | CAB learning · Pendigits | test error | Helped |
Closest papers by meaning
- Efficient Counterfactual Learning from Bandit Feedback (AAAI 2019 · similarity 0.88)
- Cross-Domain Off-Policy Evaluation and Learning for Contextual Bandits (ICLR 2025 · similarity 0.86)
- Log-Sum-Exponential Estimator for Off-Policy Evaluation and Learning (ICML 2025 · similarity 0.85)
- A General Framework for Off-Policy Learning with Partially-Observed Reward (ICLR 2025 · similarity 0.85)
- Off-policy estimation with adaptively collected data: the power of online learning (NeurIPS 2024 · similarity 0.85)
- Logarithmic Smoothing for Pessimistic Off-Policy Evaluation, Selection and Learning (NeurIPS 2024 · similarity 0.85)
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.