Research map / Probabilistic and scientific modeling / Molecules and proteins
Learning to Select Best Forecast Tasks for Clinical Outcome Prediction
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| learn auxiliary task weights with AutoSelect instead of equal-weight pretraining | Pretrain (All) · ICU Mortality, 1% data | AUC-ROC: 0.809 -> 0.833 | Helped |
| learn auxiliary task weights with AutoSelect | Pretrain (All) · ICU BP, 1% data | AUC-ROC: 0.778 -> 0.838 | Helped |
| learn auxiliary task weights with AutoSelect | Pretrain (All) · ICU KD, 1% data | AUC-ROC: 0.771 -> 0.823 | Helped |
| learn auxiliary task weights with AutoSelect | Pretrain (All) · ICU Mortality, 100% data | AUC-ROC: 0.899 -> 0.909 | No clear effect |
| pretrain with 96 forecast tasks | Supervised · ICU Mortality, 10% data | AUC-ROC: 0.853 -> 0.853 | No clear effect |
| pretrain with 96 forecast tasks | Supervised · ICU Mortality, 1% data | AUC-ROC: 0.738 -> 0.809 | Helped |
| co-train target task with all auxiliary tasks | Supervised · ICU Mortality, 1% data | AUC-ROC: 0.738 -> 0.725 | Hurt |
| pretrain only on top selected tasks (ablation) | Pretrain (All) · ICU BP, 1% data | AUC-ROC: 0.778 -> 0.812 | Helped |
| pretrain on non-top tasks only (ablation) | Pretrain (All) · ICU Mortality, 1% data | AUC-ROC: 0.809 -> 0.772 | Hurt |
| pretrain only on top selected tasks (ablation) | AutoSelect · ICU Mortality, 10% data | AUC-ROC: 0.882 -> 0.850 | Hurt |
| use mortality-learned task weights to transfer to BP | AutoSelect · ICU BP, 1% data | AUC-ROC: 0.778 -> 0.842 | Helped |
| use mortality-learned task weights to transfer to KD | AutoSelect · ICU KD, 1% data | AUC-ROC: 0.771 -> 0.833 | Helped |
Closest papers by meaning
- Contrastive Learning for Clinical Outcome Prediction with Partial Data Sources (ICML 2024 · similarity 0.81)
- Memorize and Rank: Elevating Large Language Models for Clinical Diagnosis Prediction (AAAI 2025 · similarity 0.80)
- CLR-voyance : Reinforcing Open-Ended Reasoning for Inpatient Clinical Decision Support with Outcome-Aware Rubrics (NeurIPS 2026 · similarity 0.79)
- Risk Horizons: Structured Hypothesis Spaces for Longitudinal Clinical Prediction (NeurIPS 2026 · similarity 0.79)
- Reasoning-Enhanced Healthcare Predictions with Knowledge Graph Community Retrieval (ICLR 2025 · similarity 0.78)
- GraphCare: Enhancing Healthcare Predictions with Personalized Knowledge Graphs (ICLR 2024 · similarity 0.78)
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.