Learning What Not to Impute: An Uncertainty-Aware Diffusion Framework for Meaningful Missingness
About
Missing value imputation is a fundamental task in machine learning, with most existing methods assuming that all missing entries correspond to unobserved regular values. In many real-world datasets, however, missingness may arise from two distinct sources: some entries are meaningfully missing (intrinsically absent and semantically valid), while others are missing due to the observation process and should be imputed. We formalize this distinction as a selective imputation problem, where the goal is to jointly infer which missing entries should be preserved and which should be recovered. To address this challenge, we propose Diff-Joint, a diffusion-based framework that jointly models tabular data together with a latent missingness mask. The method alternates between conditional sampling and uncertainty-aware aggregation to iteratively refine both imputed values and missingness labels. Empirical results on synthetic and real-world datasets demonstrate that Diff-Joint effectively identifies meaningfully missing entries while achieving competitive imputation accuracy and improved downstream task performance.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| CCI CHF prediction | MIMIC-IV-ED | Macro F1 Score48.08 | 20 | |
| Critical outcome prediction | MIMIC-IV-ED | Macro F190.11 | 20 | |
| Hospitalization prediction | MIMIC ED IV | Macro F149.24 | 20 | |
| ICU Transfer 12h prediction | MIMIC-IV-ED | Macro F1 Score90.99 | 20 | |
| Missing Value Imputation | Bayesian Network MCAR (test) | Accuracy78.34 | 20 | |
| Missing Value Imputation | Bayesian Network MAR (test) | Accuracy (Out-of-sample)75.32 | 20 | |
| Missing Value Imputation | Bayesian Network MNAR (test) | Accuracy (Test)72.51 | 20 | |
| Multi-class classification | Bayesian Network variable D2 | Macro-F176.55 | 20 | |
| Multi-class classification | Bayesian Network variable D3 | Macro F1 Score79.16 | 20 | |
| Data Imputation | MIMIC-IV-ED under MCAR | MAE (Output)4.59 | 20 |