Outlier Detection on Optdigits (AUC-ROC)
0.999AUC-ROCPReNet
Evaluation results
| Method | Links | |
|---|---|---|
| PReNetlabeled outliers=52025.12 | 0.999 | |
| FEAWADlabeled outliers=52025.12 | 0.994 | |
| CODlabeled outliers=52025.12 | 0.989 | |
| DevNetlabeled outliers=52025.12 | 0.988 | |
| ICAD-LLMMethod Category=Task-Specific2025.12 | 0.9788 | |
| MCMMethod Category=Task-Specific2025.12 | 0.9732 | |
| WFRDAlabeled outliers=52025.12 | 0.942 | |
| KRPDvalidation=5-fold cross-validation2025.10 | 0.926 | |
| Two-Stage LKPLO (SVM-like)validation=5-fold cross-validation2025.10 | 0.922 | |
| ICAD-LLMMethod Category=Universal2025.12 | 0.9023 | |
| Two-Stage LKPLO (Robust-Z)validation=5-fold cross-validation2025.10 | 0.888 | |
| WRPOvalidation=5-fold cross-validation2025.10 | 0.871 | |
| DeepSADlabeled outliers=52025.12 | 0.869 | |
| IForestMethod Category=Task-Specific2025.12 | 0.7973 | |
| GOADMethod Category=Task-Specific2025.12 | 0.7962 | |
| KODvalidation=5-fold cross-validation2025.10 | 0.794 | |
| IForestlabeled outliers=52025.12 | 0.739 | |
| CBLOFvalidation=5-fold cross-validation2025.10 | 0.731 | |
| IForestvalidation=5-fold cross-validation2025.10 | 0.71 | |
| DeepSVDDlabeled outliers=52025.12 | 0.667 | |
| DAGMMMethod Category=Task-Specific2025.12 | 0.6429 | |
| UniADMethod Category=Universal2025.12 | 0.6332 | |
| NeuTraL ADMethod Category=Universal2025.12 | 0.6326 | |
| ACRMethod Category=Universal2025.12 | 0.6114 | |
| REPENlabeled outliers=52025.12 | 0.609 | |
| ECODlabeled outliers=52025.12 | 0.606 | |
| LOFvalidation=5-fold cross-validation2025.10 | 0.586 | |
| KPCAvalidation=5-fold cross-validation2025.10 | 0.585 | |
| RPDvalidation=5-fold cross-validation2025.10 | 0.572 | |
| KDEvalidation=5-fold cross-validation2025.10 | 0.563 | |
| OCSVMvalidation=5-fold cross-validation2025.10 | 0.554 | |
| SODlabeled outliers=52025.12 | 0.492 | |
| KNNvalidation=5-fold cross-validation2025.10 | 0.405 |