Classification on cylinder-bands (test)
97.8AccuracyLDA
Evaluation results
| Method | Links | |
|---|---|---|
| LDAType=Euclidean, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 97.8 | |
| R-PGAType=Riemannian, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 97.8 | |
| R-RPCAType=Riemannian, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 97.8 | |
| R-ONPPType=Riemannian, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 97.2 | |
| R-LEType=Riemannian, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 96.7 | |
| R-LDAType=Riemannian, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 96.7 | |
| IsomapType=Euclidean, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 92.2 | |
| R-IsomapType=Riemannian, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 91.7 | |
| EXACTModel architecture=Linear model, Loss function=EXACT2022.05 | 77.22 | |
| HingeModel architecture=Linear model, Loss function=Hinge2022.05 | 74.44 | |
| PCAType=Euclidean, Number of components (nc)=3, Subsampling strategy=stratified, max 2000 samples2026.02 | 74.4 | |
| Cross-entropyModel architecture=Linear model, Loss function=Cross-entropy2022.05 | 72.78 | |
| SklearnModel architecture=Linear model, Loss function=Logistic regression, Solver=L-BFGS2022.05 | 72.22 |