Research map / Probabilistic and scientific modeling / Treatment effects
Efficient and Robust Transfer Learning of Optimal Individualized Treatment Regimes with Right-Censored Survival Data
Pan Zhao, Julie Josse, Shu Yang
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 augmented calibration-weighted estimator instead of IPSW | RMST value estimator · Sim scenario 1 (all models correct) | empirical bias: 0.0185 -> 0.0070 | Improved |
| use ACW instead of IPSW (coverage) | RMST value estimator · Sim scenario 1 (all models correct) | 95% CI coverage (%): 87.14 -> 94.29 | Improved |
| use calibration weights instead of logistic sampling score in IPW | IPW estimator · Sim scenario 1 (all models correct) | 95% CI coverage (%): 87.14 -> 88.29 | No clear effect |
| ignore sampling score (Naive) vs ACW | RMST value estimator · Sim scenario 1 (all models correct) | empirical bias: -0.8801 -> 0.0070 | Improved |
| use ACW instead of CW-OR when sampling model wrong but outcome correct | RMST value estimator · Sim scenario 2 (only outcome model correct) | empirical bias: -0.1312 -> 0.0055 | Hurt |
| use ACW instead of CW-OR in double robustness check (outcome model wrong, others correct) | RMST value estimator · Sim scenario 3 (only outcome model wrong) | 95% CI coverage (%): 92.29 -> 93.14 | Improved |
| use ACW when all working models are wrong | RMST value estimator · Sim scenario 4 (all models wrong) | empirical bias: 0.0115 -> 0.2644 | Hurt |
| increase sample size with ML nuisance estimation | cross-fitted ACW · Sim with ML nuisance (small vs large n) | bias | Improved |
| use ACW instead of IPW-based estimators with ML nuisance | RMST value estimator · Sim with random forests (N=6e5, m=24000) | bias/validity | Improved |
| use ACW instead of RCT-only doubly robust ITR | optimal ITR · BICAR-ICU + observational study | estimated 28-day RMST (days): 15.37 -> 19.52 | Improved |
Closest papers by meaning
- An effective framework for estimating individualized treatment rules (NeurIPS 2024 · similarity 0.88)
- Doubly Robust Fusion of Many Treatments for Policy Learning (ICML 2025 · similarity 0.88)
- Assessing the robustness of heterogeneous treatment effects in survival analysis under informative censoring (NeurIPS 2026 · similarity 0.86)
- Evaluating and Learning Optimal Dynamic Treatment Regimes under Truncation by Death (NeurIPS 2025 · similarity 0.86)
- Orthogonal Survival Learners for Estimating Heterogeneous Treatment Effects from Time-to-Event Data (NeurIPS 2025 · similarity 0.85)
- Doubly Protected Estimation for Survival Outcomes Utilizing External Controls for Randomized Clinical Trials (ICML 2025 · 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.
