NHANES
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Medical diagnostic value preservation | NHANES 2017-2018 (test) | wMAPE1.2 | 54 | |
| Aggregate Reconstruction | NHANES 8 templates (holdout) | wMAPE0.3 | 48 | |
| Classification | NHANES | Accuracy100 | 18 | |
| Conformal Prediction | nhanes | Coverage (COV)98.2 | 15 | |
| Binary Classification | NHANES (test) | AUROC85.5 | 12 | |
| Prediction under missing values | NHANES (test) | AUROC85.5 | 12 | |
| Predictive Modeling | NHANES | Accuracy89.7 | 12 | |
| Diagnosis prediction | NHANES (test) | Recall@195.94 | 11 | |
| Canonical Correlation Analysis | NHANES Race | Correlation (P_ρr)-0.0931 | 10 | |
| Canonical Correlation Analysis | NHANES Education | Correlation Coefficient (P_ρr)-0.4184 | 10 | |
| Imputation | NHANES MNAR (50% missingness) | NRMSE0.138 | 8 | |
| Imputation | NHANES MNAR (30% missingness) | NRMSE0.11 | 8 | |
| Imputation | NHANES MNAR (10% missingness) | NRMSE0.079 | 8 | |
| Health risk prediction | NHANES 2011-2014 | Angina AUROC0.7565 | 7 | |
| Ordinal Classification | nhanes | COV0.902 | 7 | |
| Imputation | NHANES 30% missingness MAR | NRMSE0.088 | 7 | |
| Imputation | NHANES 30% missingness MCAR | NRMSE0.087 | 7 | |
| Imputation | NHANES categorical variables 30% missingness (strict MAR) | Accuracy70 | 7 | |
| Health Risk Prediction | NHANES 2005–2006 (train) | Angina AUROC0.6988 | 6 | |
| AUF probability estimation | NHANES | AUF Probability59.6 | 6 | |
| Off-policy targeting | NHANES 2017-2018 | AUUC30.21 | 6 | |
| Factual Prediction | NHANES 2017-2018 | RMSE0.0285 | 6 | |
| Binary Logistic Regression | NHANES ignorable missingness | Brier Score0.194 | 6 | |
| Classification | NHANES non-ignorable missingness (test) | Brier Score0.189 | 6 | |
| Uplift Modeling | nhanes | AUROC0.78 | 6 |
Showing 25 of 35 rows