Research map / Probabilistic and scientific modeling / Treatment effects
SurvHTE-Bench: A Benchmark for Heterogeneous Treatment Effect Estimation in Survival Analysis
Shahriar Noroozizadeh, Xiaobin Shen, Jeremy Weiss, George H. Chen
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 Double-ML instead of survival meta-learners | CATE estimator · ACTG semi-synthetic | CATE RMSE: 10.65 | Improved |
| use survival meta-learner S-Learner-Survival across censoring levels | S-Learner-Survival · MIMIC-i to MIMIC-v | CATE RMSE: RMSE range 7.897--7.921 | Improved |
| use T-Learner-Survival under extreme censoring | T-Learner-Survival · MIMIC-ii (82% censoring) | CATE RMSE std: 7.980 ± 0.233 | Hurt |
| use survival meta-learners instead of outcome imputation at high censoring | S-Learner-Survival and Matching-Survival · Synthetic Scenario D | average rank: S-Learner-Survival (1.6) and Matching-Survival (2.4) | Improved |
| randomized balanced trial (RCT-50) favors outcome imputation | Double-ML and Causal Forest · Synthetic RCT-50 | average rank: Double-ML (3.60) and Causal Forest (5.60) | Improved |
| imbalanced treatment (5% treated) instead of balanced | T-Learner-Survival · Synthetic RCT-5 | average rank: drops to last place (9.00) | Hurt |
| imbalanced treatment (5% treated) instead of balanced | Double-ML · Synthetic RCT-5 | average rank: 1.80 | Improved |
| introduce positivity violation only | Causal Survival Forests · Synthetic OBS-NoPos | ranking | Hurt |
| introduce positivity violation only | Double-ML and X-Learner · Synthetic OBS-NoPos | CATE RMSE | Improved |
| add informative censoring | all methods · Synthetic InfC configurations | CATE RMSE | Hurt |
| use Margin imputation instead of other imputations | outcome imputation methods · Synthetic | imputation error | Improved |
| use DeepSurv base survival model | survival meta-learners · Synthetic | overall rank | Improved |
Closest papers by meaning
- Heterogeneous Treatment Effect in Time-to-Event Outcomes: Harnessing Censored Data with Recursively Imputed Trees (ICML 2025 · similarity 0.91)
- Orthogonal Survival Learners for Estimating Heterogeneous Treatment Effects from Time-to-Event Data (NeurIPS 2025 · similarity 0.89)
- Assessing the robustness of heterogeneous treatment effects in survival analysis under informative censoring (NeurIPS 2026 · similarity 0.89)
- Do Contemporary Causal Inference Models Capture Real-World Heterogeneity? Findings from a Large-Scale Benchmark (ICLR 2025 · similarity 0.88)
- Doubly robust identification of treatment effects from multiple environments (ICLR 2025 · similarity 0.87)
- Using Surrogates in Covariate-adjusted Response-adaptive Randomization Experiments with Delayed Outcomes (NeurIPS 2024 · similarity 0.87)
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.
