Ordinal Classification on APTOS 2019 (val)
78.9AccCE-E
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CE-EBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 78.9 | 82.3 | 1.043 | 0.31 | 82.4 | |
| OCEBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 78.4 | 89.7 | 1.081 | 0.324 | 80.3 | |
| CE-βBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 78.2 | 85.4 | 1.095 | 0.329 | 80.4 | |
| CEBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 78.1 | 89.6 | 1.098 | 0.33 | 79.8 | |
| CE-BBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 77.4 | 78.4 | 1.096 | 0.322 | 82.3 | |
| CEBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 77.4 | 87.8 | 1.135 | 0.341 | 79.5 | |
| OCEBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 77.2 | 87.8 | 1.147 | 0.345 | 79.1 | |
| CE-βBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 76.4 | 82.8 | 1.174 | 0.351 | 79.3 | |
| CE-EBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 76.2 | 80.3 | 1.164 | 0.344 | 80.7 | |
| CE-βBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75.9 | 85.9 | 1.268 | 0.393 | 72 | |
| CE-EBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75.9 | 84.5 | 1.249 | 0.383 | 73.2 | |
| OCEBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75.8 | 82.1 | 1.22 | 0.368 | 77.4 | |
| CE-PBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75.6 | 86 | 1.219 | 0.366 | 76.6 | |
| CEBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75.6 | 82.7 | 1.244 | 0.378 | 77 | |
| CE-PBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75.4 | 76.2 | 1.266 | 0.387 | 72.9 | |
| CE-BBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 75 | 80.7 | 1.292 | 0.396 | 73.1 | |
| CE-PBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 74.3 | 82.6 | 1.307 | 0.396 | 76.7 | |
| OLBackbone=DenseNet121, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 74 | 71.4 | 1.26 | 0.37 | 77.8 | |
| CE-BBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 73.9 | 69.4 | 1.263 | 0.371 | 80.2 | |
| OLBackbone=InceptionV3, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 73.4 | 66.2 | 1.307 | 0.387 | 76.1 | |
| OLBackbone=VGG19, Cost Matrix Type=Symmetric, Evaluation Protocol=5-fold Cross-Validation2026.06 | 72.6 | 67.7 | 1.366 | 0.409 | 73.5 |