Conditional Anomaly Detection on Auto MPG UCI ML (2/3, 1/3 train-test split)
72.6Mean Anomaly Agreement Score (AUC)SoftHAD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SoftHADNumber of runs=100, Train-test split=2/3, 1/32026.04 | 72.6 | 1.7 | |
| SoftHAD2026.05 | 72.6 | 1.7 | |
| QDANumber of runs=100, Train-test split=2/3, 1/32026.04 | 65.9 | 2.9 | |
| QDA2026.05 | 65.9 | 2.9 | |
| wk–NNNumber of runs=100, Train-test split=2/3, 1/32026.04 | 61.4 | 2.3 | |
| weighted k-NN2026.05 | 61.4 | 2.3 | |
| 1-class SVMNumber of runs=100, Train-test split=2/3, 1/3, Kernel type=RBF2026.04 | 50.1 | 3.5 | |
| one-class SVM2026.05 | 50.1 | 3.5 | |
| SVMNumber of runs=100, Train-test split=2/3, 1/3, Kernel type=RBF2026.04 | 37.1 | 8.6 | |
| SVM2026.05 | 37.1 | 8.6 |