Three-class Classification (CN/MCI/AD) on ADNI 129-subject (test)
67.44AccuracySurface-based Multimodal Framework
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Surface-based Multimodal FrameworkT=3002026.06 | 67.44 | 69.02 | 66.22 | 67.09 | 83.56 | 0.3982 | |
| TabPFNInput=Original Data2026.06 | 65.89 | 64.24 | 60.13 | 61.13 | 78.63 | 0.3618 | |
| XGBoost2026.06 | 62.79 | 62.04 | 59.78 | 59.97 | 72.59 | 0.3035 | |
| AutoGluon2026.06 | 62.02 | 60.8 | 51.53 | 54.05 | 73.52 | 0.2755 | |
| LightGBM2026.06 | 59.69 | 56.95 | 57.31 | 56.48 | 70.7 | 0.2486 | |
| MulT2026.06 | 59.69 | 54.87 | 59.54 | 56.63 | 75.26 | 0.2905 | |
| TabPFNInput=Tabular only2026.06 | 58.14 | 68.3 | 50.2 | 54.64 | 75.37 | 0.2141 | |
| Random Forest2026.06 | 55.81 | 58.49 | 44.81 | 47.67 | 66.34 | 0.1465 | |
| TabNet2026.06 | 52.71 | 29.09 | 33.28 | 29.81 | 56.67 | 0.011 |