Research map / Reinforcement learning / Policy gradient and actor-critic
Asymptotic Inference for Multi-Stage Stationary Treatment Policy with Variable Selection
Daiqi Gao, Yufeng Liu, Donglin Zeng
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 Q-function-based nuisance Q(1) instead of Q(0) or Q(2) in the doubly robust estimator (ablation) | MSTP · Simulation Scenarios 1 and 2 | average reward and MAD | Improved |
| Q(2) depends on importance sampling weights, used with larger T (ablation) | MSTP · Simulation, T up to 10 | numerical stability | Hurt |
| estimate a stochastic MSTP policy rather than PEARL's deterministic policy | MSTP vs PEARL · Simulation Scenario 1, T=1 | average reward | Hurt |
| MSTP stochastic policy versus PEARL | MSTP vs PEARL · Simulation Scenario 2, T=1 | average reward | Improved |
| add one-step correction to the sparse estimator for inference (ablation) | MSTP confidence intervals · Simulation, T=10 | coverage probability | Improved |
| increase horizon T | MSTP inference · Simulation | coverage probability | Hurt |
| use Q(1) nuisance versus Q(0), Q(2) and PEARL for one-step estimator (ablation) | MSTP one-step estimator · Simulation | MAD and coverage | Improved |
| learn policy on OhioT1DM versus behavior policy | MSTP · OhioT1DM | average reward: -6.260 -> -2.367 | Improved |
Closest papers by meaning
- Semi-Supervised Off-Policy Reinforcement Learning and Value Estimation for Dynamic Treatment Regimes (JMLR 2023 · similarity 0.87)
- Estimating Continuous Treatment Effects with Two-Stage Kernel Ridge Regression (ICML 2026 · similarity 0.87)
- Continuous Treatment Effect Estimation Using Gradient Interpolation and Kernel Smoothing (AAAI 2024 · similarity 0.85)
- Enhancing Treatment Effect Estimation via Active Learning: A Counterfactual Covering Perspective (ICML 2025 · similarity 0.84)
- Learning Shadow Variable Representation for Treatment Effect Estimation under Collider Bias (ICML 2024 · similarity 0.84)
- Distributionally Robust Policy Evaluation and Learning for Continuous Treatment with Observational Data (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.
