Digit Classification on SVHN -> MNIST
0.944Accuracydistributionally robust learning framework
Evaluation results
| Method | Links | |
|---|---|---|
| distributionally robust learning frameworkBase classifier=STAR, Hyperparameter selection=Validation-based, Target sample size=10^22026.01 | 0.944 | |
| distributionally robust learning frameworkBase classifier=STAR, Hyperparameter selection=LODO-CV, Target sample size=10^22026.01 | 0.929 | |
| distributionally robust learning frameworkBase classifier=CDAN, Hyperparameter selection=Validation-based, Target sample size=10^22026.01 | 0.925 | |
| Generate To AdaptBackbone=Modified LeNet2017.04 | 0.924 | |
| distributionally robust learning frameworkBase classifier=ERM, Hyperparameter selection=Validation-based, Target sample size=10^22026.01 | 0.92 | |
| distributionally robust learning frameworkBase classifier=STAR, Hyperparameter selection=Validation-based, Target sample size=102026.01 | 0.913 | |
| distributionally robust learning frameworkBase classifier=ERM, Hyperparameter selection=Validation-based, Target sample size=102026.01 | 0.87 | |
| distributionally robust learning frameworkBase classifier=CDAN, Hyperparameter selection=Validation-based, Target sample size=102026.01 | 0.87 | |
| distributionally robust learning frameworkBase classifier=STAR, Hyperparameter selection=LODO-CV, Target sample size=102026.01 | 0.851 | |
| ATDOCTarget sample size=10^22026.01 | 0.833 | |
| DRCNBackbone=Modified LeNet2017.04 | 0.82 | |
| MCDTarget sample size=10^22026.01 | 0.791 | |
| STARTarget sample size=10^22026.01 | 0.764 | |
| ADDABackbone=Modified LeNet2017.04 | 0.76 | |
| CORALTarget sample size=10^22026.01 | 0.754 | |
| RevGradBackbone=Modified LeNet2017.04 | 0.739 | |
| STARTarget sample size=102026.01 | 0.668 | |
| DANNTarget sample size=10^22026.01 | 0.66 | |
| CORALTarget sample size=102026.01 | 0.636 | |
| CDANTarget sample size=10^22026.01 | 0.634 | |
| MCDTarget sample size=102026.01 | 0.613 | |
| DANNTarget sample size=102026.01 | 0.612 | |
| Source onlyBackbone=Modified LeNet2017.04 | 0.603 | |
| ERM (Src-only)Target sample size=None (Source-only)2026.01 | 0.596 | |
| ATDOCTarget sample size=102026.01 | 0.591 | |
| CDANTarget sample size=102026.01 | 0.569 | |
| MK-MMDTarget sample size=10^22026.01 | 0.5 | |
| MK-MMDTarget sample size=102026.01 | 0.483 |