Binary Classification on MNIST 0vsAll (test)
100AccuracyDropConnect
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DropConnectProtocol=one-vs-all2026.06 | 100 | 0.01 | 0 | |
| LikelihoodProtocol=one-vs-all2026.06 | 100 | 0.01 | 0.14 | |
| EK-NNProtocol=one-vs-all2026.06 | 100 | 0.03 | 0.04 | |
| MC-DropoutProtocol=one-vs-all2026.06 | 100 | 0.01 | 0 | |
| Deep EnsembleProtocol=one-vs-all2026.06 | 100 | 0.01 | 0 | |
| Bayesian DeepProtocol=one-vs-all2026.06 | 100 | 0.01 | 0 | |
| Bayesian Log. Reg.Protocol=one-vs-all2026.06 | 99 | 0.01 | 0 | |
| Random ForestProtocol=one-vs-all2026.06 | 99 | 0.04 | 0.01 | |
| CentroidsProtocol=one-vs-all2026.06 | 94 | 0.35 | 0.07 |