Research map / Probabilistic and scientific modeling / Treatment effects
Near-optimal Individualized Treatment Recommendations
Haomiao Meng, Ying-Qi Zhao, Haoda Fu, Xingye Qiao
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 |
|---|---|---|---|
| recommend a near-optimal set (A-ITR) instead of a single treatment | One-step OWL · Simulation Example 1 (p=5) | empirical weighted outcome: 2.23 -> 2.19 | Improved |
| recommend a near-optimal set (A-ITR) instead of a single treatment | Regression · Simulation Example 2 (p=5) | empirical weighted outcome: 1.60 -> 1.45 | Improved |
| recommend a near-optimal set (A-ITR) instead of a single treatment | One-step Gaussian · T2DM data | cross-validated weighted outcome: 0.939 -> 0.935 | Improved |
| recommend a near-optimal set (A-ITR) instead of a single treatment | One-step Linear · T2DM data | cross-validated weighted outcome: 1.150 -> 1.033 | Improved |
| use one-step OWL with bent hinge loss instead of regression-based plug-in | A-ITR · Simulation Example 1 (p=5) | empirical weighted outcome: 2.37 -> 2.19 | Improved |
| use one-step OWL instead of two-step OWL | A-ITR · Simulation Example 2 (p=10) | empirical weighted outcome: 1.35 -> 1.31 | Improved |
| use one-step OWL instead of two-step OWL | A-ITR · Simulation Example 3 (p=10) | empirical weighted outcome: 3.32 -> 3.33 | Mixed |
| use regression-based plug-in A-ITR instead of classification-based | Regression A-ITR · Simulation Example 1 (p=5) | interval length (R1): (1.98, 3.09) vs (1.99, 2.23) | Hurt |
| use Gaussian kernel instead of linear kernel (ablation) | One-step OWL A-ITR · T2DM data | cross-validated weighted outcome: 1.033 -> 0.935 | Improved |
| use Gaussian kernel instead of linear kernel (ablation) | Two-step A-ITR · T2DM data | cross-validated weighted outcome: 0.975 -> 0.947 | Improved |
| increase covariate dimension from p=5 to p=10 (ablation) | One-step A-ITR · Simulation Example 1 | empirical weighted outcome: 2.19 -> 2.30 | Hurt |
Closest papers by meaning
- An effective framework for estimating individualized treatment rules (NeurIPS 2024 · similarity 0.88)
- Efficient and Robust Transfer Learning of Optimal Individualized Treatment Regimes with Right-Censored Survival Data (JMLR 2025 · similarity 0.87)
- Doubly Robust Fusion of Many Treatments for Policy Learning (ICML 2025 · similarity 0.86)
- PairNet: Training with Observed Pairs to Estimate Individual Treatment Effect (ICML 2024 · similarity 0.82)
- SAFER: A Calibrated Risk-Aware Multimodal Recommendation Model for Dynamic Treatment Regimes (ICML 2025 · similarity 0.82)
- Defining Expertise: Applications to Treatment Effect Estimation (ICLR 2024 · similarity 0.81)
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.
