Research map / Probabilistic and scientific modeling / Treatment effects
Gradient Regularized V-Learning for Dynamic Treatment Regimes
Yao Zhang, Mihaela van der Schaar
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 |
|---|---|---|---|
| GRV estimator vs AIPW for value estimation at n=1000 (treatment cost) | GRV · Treatment cost trade-off | MSE: .139 ˘ .006 -> .134 ˘ .004 | Improved |
| GRV vs AIPW at n=10000 | GRV · Treatment cost trade-off | MSE: .023 ˘ .003 vs .022 ˘ .002 | No clear effect |
| GRV vs AIPW at n=1000 (survival) | GRV · Survival rate maximization | MSE: .047 ˘ .005 -> .032 ˘ .002 | Improved |
| GRV vs IPW at n=1000 (survival) | GRV · Survival rate maximization | MSE: .332 ˘ .035 -> .032 ˘ .002 | Improved |
| Increase training sample size 1000 to 10000 | GRV · Treatment cost trade-off | MSE: .134 ˘ .004 -> .022 ˘ .002 | Improved |
| Learn DTR with GRV-B instead of IPW/BQL | GRV-B · Treatment cost trade-off n=1000 | DTR value: .748 ˘ .002 -> .761 ˘ .003 | Improved |
| Learn DTR with GRV-B vs BQL (survival) | GRV-B · Survival rate maximization n=5000 | DTR value: .198 ˘ .003 -> .224 ˘ .002 | Improved |
| Optimize all treatment rules together (GRV-S) instead of backward (GRV-B) (ablation) | GRV · Treatment cost trade-off n=1000 | DTR value: .761 ˘ .003 -> .753 ˘ .004 | Hurt |
| GRV-S vs GRV-B in survival n=1000 (ablation) | GRV · Survival rate maximization n=1000 | DTR value: .182 ˘ .004 -> .146 ˘ .005 | Hurt |
| Ablate the gradient regularizer (ablation) | GRV · Appendix simulations | performance | Improved |
| GRV-S vs BQL, survival n=10000 | GRV-S · Survival rate maximization n=10000 | DTR value: .215 ˘ .002 vs .213 ˘ .002 | No clear effect |
Closest papers by meaning
- Evaluating and Learning Optimal Dynamic Treatment Regimes under Truncation by Death (NeurIPS 2025 · similarity 0.85)
- Continuous Treatment Effect Estimation Using Gradient Interpolation and Kernel Smoothing (AAAI 2024 · similarity 0.85)
- Model-agnostic meta-learners for estimating heterogeneous treatment effects over time (ICLR 2025 · similarity 0.85)
- Orthogonal Survival Learners for Estimating Heterogeneous Treatment Effects from Time-to-Event Data (NeurIPS 2025 · similarity 0.85)
- Doubly Robust Fusion of Many Treatments for Policy Learning (ICML 2025 · similarity 0.84)
- ODE Discovery for Longitudinal Heterogeneous Treatment Effects Inference (ICLR 2024 · 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.
