Multiple Instance Learning Classification on MUSK1
98.2AccuracyTG-RGMIL [0,1,2]D
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TG-RGMIL [0,1,2]DTopological features=0D, 1D, 2D2023.07 | 98.2 | — | |
| ASMIL2026.03 | 97.1 | — | |
| RGMIL2026.03 | 96.8 | — | |
| PSMIL2026.03 | 96.8 | — | |
| GAPMILwD (2025)Reference Year=20252023.07 | 96.7 | — | |
| TG-RGMIL [0,1]DTopological features=0D, 1D2023.07 | 96.6 | — | |
| APMILwD (2025)Reference Year=20252023.07 | 96.4 | — | |
| DEMIL2026.03 | 96.3 | — | |
| Snuffy2024.08 | 96.1 | 98.9 | |
| DSMIL2026.03 | 95.9 | — | |
| PPPM-kernel2016.10 | 95.6 | — | |
| TG-RGMIL [0]DTopological features=0D2023.07 | 94.6 | — | |
| RGMIL (2023)Reference Year=20232023.07 | 94 | — | |
| MI-ANFIS2016.10 | 93.49 | — | |
| DSMIL (2021)Reference Year=20212023.07 | 93.2 | — | |
| Bagging-APR2016.10 | 92.8 | — | |
| TransMIL2026.03 | 92.7 | — | |
| BDRMIL (2022)Reference Year=20222023.07 | 92.6 | — | |
| APR2016.10 | 92.4 | — | |
| Citation-KNN2016.10 | 92.4 | — | |
| DistNet (2023)Reference Year=20232023.07 | 92.3 | — | |
| GNNMIL (2019)Reference Year=20192023.07 | 91.7 | — | |
| ABMIL2026.03 | 91.6 | — | |
| RBF-MIP2016.10 | 91.3 | — | |
| DPMIL (2018)Reference Year=20182023.07 | 90.7 | — | |
| RBF-Bag-Unit2016.10 | 90.3 | — | |
| GAPMIL (2018)Reference Year=20182023.07 | 90 | — | |
| MIGraph2016.10 | 90 | — | |
| APMIL (2018)Reference Year=20182023.07 | 89.2 | — | |
| DD2016.10 | 88.9 | — | |
| miGraph2016.10 | 88.9 | — | |
| MI-NN2016.10 | 88 | — | |
| MI-kernel2016.10 | 88 | — | |
| mi-SVM2016.10 | 87.4 | — | |
| MILES2016.10 | 86.3 | — | |
| ALP-SVM2016.10 | 86.3 | — | |
| DD-SVM2016.10 | 85.8 | — | |
| MIForest2016.10 | 85 | — | |
| EM-DD2016.10 | 84.8 | — | |
| BP-MIP2016.10 | 83.7 | — | |
| ABMIL-Gated2024.08 | 83.1 | 83.7 | |
| ABMIL2024.08 | 82.6 | 82.4 | |
| Mean-pooling2024.08 | 80 | 86.9 | |
| DSMIL2024.08 | 78.6 | 85.2 | |
| MI-SVM2016.10 | 77.9 | — | |
| Max-pooling2024.08 | 72.8 | 79.9 | |
| ABMIL2026.06 | — | 85.7 | |
| Fact. (cont, MLP)2026.06 | — | 93.9 | |
| Fact. (cont, tree)2026.06 | — | 90.3 | |
| Fact. (disc, lookup)2026.06 | — | 87 | |
| Fact. (disc, MLP)2026.06 | — | 76.7 | |
| Fact. (disc, tree)2026.06 | — | 40.6 | |
| ICMILtraining steps=40,000, batch size=128, embedding dimension=256, MLP hidden size=10542026.06 | — | 93.3 | |
| Joint2026.06 | — | 91 | |
| MeanLogReg2026.06 | — | 91.6 | |
| MixedJoint weight=0.70, Fact. MLP weight=0.15, Fact. lookup weight=0.152026.06 | — | 92.5 | |
| SVM-Summ2026.06 | — | 94.7 | |
| TabPFN-Cluster2026.06 | — | 88 | |
| TabPFN-Concat2026.06 | — | 91.1 | |
| TabPFN-Subsample2026.06 | — | 92.5 |