AI Paper Atlas

Research map / Reinforcement learning / Offline RL

Distributionally Robust Policy Evaluation and Learning in Offline Contextual Bandits

Nian Si, Fan Zhang, Zhengyuan Zhou, Jose Blanchet

ICML 2020 · Reinforcement learning › Offline RL · approach: Diffusion models for offline decision making · 14 citations (OpenAlex)

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 changedWhereMeasuredEffect
Normalizing the DR objective by the propensity sum (stable version, Algorithm 3) instead of unnormalized (Algorithm 2) (ablation)DR policy learning with depth-3 trees · Simulation worst-case KL ball test setQ_full_DRO value: 0.075 vs 0.067Improved
Normalizing the DR objective (stable version) in policy learning (ablation)DR policy learning · Simulation same-environment test setQ_full_emp value: 0.243 vs 0.240No clear effect
Distributionally robust objective (stable DRO policy learning) vs non-robust empirical policy learningDRO policy learning · Simulation worst-case KL ball test setQ_full_DRO value: 0.075 vs 0.059Improved
Distributionally robust objective (stable DRO) vs non-robust policy learningDRO policy learning · Simulation min over 100 KL-perturbed test setsQ_full_min value: 0.203 vs 0.191Improved
Distributionally robust objective (stable DRO) vs non-robust policy learning in unperturbed environmentDRO policy learning · Simulation same-environment test setQ_full_emp value: 0.243 vs 0.243No clear effect
Stable (normalized) estimator vs unnormalized estimator for fixed-policy evaluation (ablation)DR policy evaluation · Simulation n=10^5, delta=0.2policy value estimate: 0.0245 vs 0.0247No clear effect
Increasing sample size from 10^3 to 10^5 for DR policy evaluationDR policy evaluation (Algorithm 1) · Simulation delta=0.2standard error of estimate: ±.0312 at n=10^3 to ±.0030 at n=10^5Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy