Tabular Imputation on Miss-Bench 42 OpenML datasets small tables v1
0.904MCAR ScoreTabImpute
Evaluation results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| TabImputeZero-shot imputation=true, Small-table regime=<= 170 rows, Standardized RMSE at the column level=1 - Normalized RMSE2025.10 | 0.904 | 0.943 | 0.775 | 0.921 | 0.937 | 0.93 | 0.955 | 0.825 | 0.805 | 0.923 | 0.933 | 0.914 | 0.766 | 0.887 | |
| OTSmall-table regime=<= 170 rows, Standardized RMSE at the column level=1 - Normalized RMSE2025.10 | 0.866 | 0.873 | 0.693 | 0.788 | 0.857 | 0.853 | 0.918 | 0.787 | 0.773 | 0.87 | 0.868 | 0.822 | 0.739 | 0.824 | |
| MissForestSmall-table regime=<= 170 rows, Standardized RMSE at the column level=1 - Normalized RMSE2025.10 | 0.854 | 0.838 | 0.7 | 0.83 | 0.85 | 0.802 | 0.897 | 0.58 | 0.625 | 0.835 | 0.848 | 0.881 | 0.749 | 0.791 | |
| HyperImputeSmall-table regime=<= 170 rows, Standardized RMSE at the column level=1 - Normalized RMSE2025.10 | 0.814 | 0.785 | 0.751 | 0.868 | 0.854 | 0.837 | 0.883 | 0.631 | 0.669 | 0.783 | 0.844 | 0.88 | 0.81 | 0.801 | |
| K-Nearest NeighborsSmall-table regime=<= 170 rows, Standardized RMSE at the column level=1 - Normalized RMSE2025.10 | 0.772 | 0.8 | 0.68 | 0.872 | 0.893 | 0.851 | 0.94 | 0.658 | 0.635 | 0.81 | 0.807 | 0.905 | 0.561 | 0.783 |