Semi-supervised learning on MNIST 1,000 labels
95AccuracyAEVB-IAF
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| AEVB-IAFPrior=Pinwheel, Classifier=KNN (K=20)2024.01 | 95 | 2.1 | 1.06 | |
| DDVIPrior=Pinwheel, Classifier=KNN (K=20)2024.01 | 95 | 0.24 | 1.06 | |
| AEVBPrior=Pinwheel, Classifier=KNN (K=20)2024.01 | 93 | 11.15 | 1.36 | |
| DDVIPrior=Swiss Roll, Classifier=KNN (K=20)2024.01 | 92 | 2.89 | 2.09 | |
| AEVB-IAFPrior=Square, Classifier=KNN (K=20)2024.01 | 91 | 2.67 | 0.9 | |
| AEVB-IAFPrior=Swiss Roll, Classifier=KNN (K=20)2024.01 | 90 | 5.38 | 2.75 | |
| DDVIPrior=Square, Classifier=KNN (K=20)2024.01 | 90 | 0.02 | 1.49 | |
| AAEBPrior=Pinwheel, Classifier=KNN (K=20)2024.01 | 89 | — | 1.55 | |
| AAEBPrior=Swiss Roll, Classifier=KNN (K=20)2024.01 | 88 | — | 3.07 | |
| AEVBPrior=Square, Classifier=KNN (K=20)2024.01 | 86 | 10.26 | 1.68 | |
| AAEBPrior=Square, Classifier=KNN (K=20)2024.01 | 82 | — | 1.48 | |
| AEVBPrior=Swiss Roll, Classifier=KNN (K=20)2024.01 | 68 | 15.29 | 4.6 |