Semi-supervised Image Classification on MNIST (test)
0.0084Test ErrorLadder
Evaluation results
| Method | Links | |
|---|---|---|
| LadderNumber of labeled samples=10002017.04 | 0.0084 | |
| Auxiliary Deep Generative Model (ADGM)Number of labels=1002016.02 | 0.0096 | |
| Ladder NetworkNumber of labels=1002016.02 | 0.0106 | |
| LadderNumber of labeled samples=1002017.04 | 0.0106 | |
| Skip Deep Generative Model (SDGM)Number of labels=1002016.02 | 0.0132 | |
| DEPICTNumber of labeled samples=30002017.04 | 0.0191 | |
| SWWAENumber of labeled samples=30002017.04 | 0.021 | |
| DEPICTNumber of labeled samples=10002017.04 | 0.021 | |
| VATNumber of labels=1002016.02 | 0.0212 | |
| M1+M2Number of labeled samples=30002017.04 | 0.0218 | |
| M1+M2Number of labeled samples=10002017.04 | 0.024 | |
| MTCNumber of labeled samples=30002017.04 | 0.0257 | |
| DEPICTNumber of labeled samples=1002017.04 | 0.0265 | |
| PL-DAENumber of labeled samples=30002017.04 | 0.0269 | |
| SWWAENumber of labeled samples=10002017.04 | 0.0283 | |
| CAENumber of labeled samples=30002017.04 | 0.0322 | |
| M1+M2Number of labels=1002016.02 | 0.0333 | |
| M1+M2Number of labeled samples=1002017.04 | 0.0333 | |
| T-SVMNumber of labeled samples=30002017.04 | 0.0345 | |
| PL-DAENumber of labeled samples=10002017.04 | 0.0346 | |
| MTCNumber of labeled samples=10002017.04 | 0.0364 | |
| AtlasRBFNumber of labeled samples=10002017.04 | 0.0368 | |
| CAENumber of labeled samples=10002017.04 | 0.0477 | |
| T-SVMNumber of labeled samples=10002017.04 | 0.0538 | |
| AtlasRBFNumber of labeled samples=1002017.04 | 0.081 | |
| SWWAENumber of labeled samples=1002017.04 | 0.0871 | |
| PL-DAENumber of labeled samples=1002017.04 | 0.1049 | |
| M1+TSVMNumber of labels=1002016.02 | 0.1182 | |
| MTCNumber of labeled samples=1002017.04 | 0.1203 | |
| CAENumber of labeled samples=1002017.04 | 0.1347 | |
| T-SVMNumber of labeled samples=1002017.04 | 0.1681 |