Image Classification on MNIST (0-4) (test)
99.91AccuracyMC-Dropout
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MC-Dropout2022.01 | 99.91 | 99.62 | 0.83 | |
| Deep-Ensembles2022.01 | 99.89 | 99.74 | 0.15 | |
| Max. Softmax2022.01 | 99.87 | 99.68 | 0.11 | |
| Entropy2022.01 | 99.87 | 99.66 | 0.11 | |
| One-vs-All Baseline2022.01 | 99.84 | 99.84 | 0.12 | |
| Confident Classifier2022.01 | 99.82 | 99.62 | 0.16 | |
| UQGANMC-Dropout=true2022.01 | 99.8 | 99.42 | 1.38 | |
| Entropy OracleOracle Evaluation=true2022.01 | 99.79 | 98.93 | 0.82 | |
| One-vs-All OracleOracle Evaluation=true2022.01 | 99.77 | 99.47 | 0.14 | |
| UQGAN2022.01 | 99.74 | 99.35 | 0.15 | |
| GEN2022.01 | 99.7 | 98.13 | 1.98 | |
| Bayes-by-Backprop2022.01 | 99.67 | 99.5 | 0.78 |