AI Paper Atlas

Research map / Reinforcement learning / Offline RL

Semi-Supervised Off-Policy Reinforcement Learning and Value Estimation for Dynamic Treatment Regimes

Aaron Sonabend-W, Nilanjana Laha, Ashwin N. Ananthakrishnan, Tianxi Cai, Rajarshi Mukherjee

JMLR 2023 · Reinforcement learning › Offline RL · approach: Behavior regularization for offline RL

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
add unlabeled data via semi-supervised imputationsupervised Q-learning · Simulation gamma11 (n=500, N=10000)empirical standard error: 0.22 -> 0.12Improved
add unlabeled data via semi-supervised imputationsupervised Q-learning · Simulation gamma12 (n=500, N=10000)empirical standard error: 0.29 -> 0.17Improved
use basis expansion instead of random forest imputationSSL Q-learning · Simulation gamma21 (n=500, N=10000)relative efficiency: 1.80 -> 1.86Improved
add unlabeled data to the doubly robust value estimatorsupervised DR value estimator · EHR setting, correct modelsrelative efficiency: 1.55Improved
add unlabeled data to the doubly robust value estimatorsupervised DR value estimator · EHR setting, misspecified Qempirical standard error: 0.07 -> 0.04Improved
add unlabeled data to the doubly robust value estimatorsupervised DR value estimator · Continuous setting, misspecified propensityempirical standard error: 0.14 -> 0.11Improved
add unlabeled data to the value estimatorsupervised DR value estimator · IBD datastandard error: 0.486 -> 0.397Improved
add unlabeled data to the value estimatorsupervised DR value estimator · IBD datavalue estimate: 0.851 -> 0.871Improved
misspecify the Q function modelSSL and supervised estimators · Simulation interaction parametersbiasNo clear effect
use semi-supervised Q-learningsupervised Q-learning · IBD datapower to detect significant predictorsImproved

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