Image Classification on Shift-MNIST (test)
93.43AccuracyGIW
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GIWShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 93.43 | — | |
| BNN Gaussian PriorBackbone=MLP2021.06 | 90.83 | -0.598 | |
| Deep EnsembleBackbone=MLP2021.06 | 88.73 | -0.527 | |
| MAPBackbone=MLP2021.06 | 88.7 | -0.527 | |
| BNN EmpCov PriorBackbone=MLP2021.06 | 86.95 | -1.146 | |
| Pretrain-valShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 86.43 | — | |
| ReweightShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 80.99 | — | |
| R-DIWShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 80.15 | — | |
| MW-NetShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 79.76 | — | |
| DIWShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 79.73 | — | |
| Val-onlyShift type=Class-prior Shift, Shift intensity=10, Case=iii, Trials=52023.05 | 78.94 | — | |
| Deep EnsembleBackbone=CNN2021.06 | 72.99 | -1.041 | |
| BNN EmpCov PriorBackbone=CNN2021.06 | 64.41 | -1.45 | |
| BNN Gaussian PriorBackbone=CNN2021.06 | 64.27 | -1.326 | |
| MAPBackbone=CNN2021.06 | 48.63 | -2.206 |