Research map / Probabilistic and scientific modeling / Treatment effects
Debiased Causal Tree: Heterogeneous Treatment Effects Estimation with Unmeasured Confounding
Caizhi Tang, Huiyuan Wang, Xinyu Li, Qing Cui, Ya-Lin Zhang, Feng Zhu, Longfei Li, JUN ZHOU, Linbo Jiang
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 |
|---|---|---|---|
| add confounding entropy to the de-bias loss (GBCT vs GBCT-ND) (ablation) | GBCT · Simulation Scenario I, phi=1 | MAE: 3.89 ± 0.57 -> 1.71 ± 0.06 | Improved |
| add confounding entropy (GBCT vs GBCT-ND) (ablation) | GBCT · Simulation Scenario II, phi=1 | MAE | Improved |
| use full confounding entropy instead of only first term (GBCT vs GBCT-B) (ablation) | GBCT · Simulation Scenario I, phi=1 | MAE: 2.85 ± 0.34 -> 1.71 ± 0.06 | Improved |
| use full confounding entropy instead of only first term (GBCT vs GBCT-B) (ablation) | GBCT · Simulation Scenario I, phi=0.2 | MAE: 1.06 ± 0.01 -> 1.05 ± 0.01 | No clear effect |
| GBCT vs XLearner at low imbalance | GBCT · Simulation Scenario I, phi=0.5 | MAE: 1.19 ± 0.10 -> 1.25 ± 0.03 | Hurt |
| GBCT vs GRF as imbalance increases | GBCT · Simulation Scenario I, phi=1 | MAE: 1.96 ± 0.20 -> 1.71 ± 0.06 | Improved |
| GBCT vs GRF under unmeasured confounding | GBCT · Simulation Scenario II, phi=1 | MAE: 2.38 ± 0.17 -> 1.81 ± 0.10 | Improved |
| GBCT vs TLearner on biased observational data | GBCT · Credit card, medium-risk, 6k | MAE_ATE: 178.6 ± 116.4 -> 61.5 ± 29.5 | Improved |
| GBCT vs DR-RF on biased observational data | GBCT · Credit card, medium-risk, 6k | MAE_ATE: 408.6 ± 9.4 -> 61.5 ± 29.5 | Improved |
| GBCT vs DML-RF on biased observational data | GBCT · Credit card, medium-risk, 6k | MAE_ATE: > 999.9 -> 61.5 ± 29.5 | Improved |
| GBCT vs DR-RF on RCT data | GBCT · Credit card, medium-risk, 2k, RCT | MAE_ATE: 8.6 ± 7.5 -> 13.2 ± 7.6 | Hurt |
| GBCT vs SLearner on biased data | GBCT · Credit card, low-risk, 6k | MAE_ATE: 45.1 ± 39.6 -> 19.3 ± 22.5 | Improved |
Closest papers by meaning
- Automating the Selection of Proxy Variables of Unmeasured Confounders (ICML 2024 · similarity 0.90)
- A Non-parametric Direct Learning Approach to Heterogeneous Treatment Effect Estimation under Unmeasured Confounding (NeurIPS 2024 · similarity 0.90)
- Detecting and Measuring Confounding Using Causal Mechanism Shifts (NeurIPS 2024 · similarity 0.89)
- Doubly robust identification of treatment effects from multiple environments (ICLR 2025 · similarity 0.89)
- Data Fusion for Partial Identification of Causal Effects (NeurIPS 2025 · similarity 0.89)
- Do Contemporary Causal Inference Models Capture Real-World Heterogeneity? Findings from a Large-Scale Benchmark (ICLR 2025 · similarity 0.89)
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.
