Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

ADBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Anomaly DetectionADBench noisy-context scenario 1.0
F1 Score84.78
532
Anomaly DetectionADBench Tabular (aggregated across 47 datasets)
Average AUROC86.0246
41
Anomaly DetectionADBench
Mean AUCROC90.43
34
Anomaly DetectionADBench MNIST-C
F1 Score21.86
26
Anomaly DetectionADBench breastw
F1 Score94.15
26
Anomaly DetectionADBench aloi
Average AUC PR10.36
26
Outlier DetectionADBench v1 (train)
ALOI0.147
19
Anomaly DetectionADBench 57 Datasets (test)
AP54.7
19
Anomaly DetectionADBench ID 2
AUCROC92.68
17
Anomaly DetectionADBench ID 1
AUCROC66.95
17
Anomaly DetectionADBench anomaly ratio categories
AP (< 3%)31.3
15
Anomaly DetectionADBench unsupervised (contaminated) setting
AUC-PR34.3
15
Anomaly DetectionADBench 38 real-world datasets (clean-context)
F1 Score65.52
14
Anomaly DetectionADBench All Datasets 57 datasets
ROC-AUC81.49
10
Anomaly DetectionADBench Low Dimensionality, d ≤ 20 (24 datasets)
ROC-AUC90.14
10
Anomaly DetectionADBench Medium Dimensionality (20 < d ≤ 100)
ROC-AUC80.13
10
Anomaly DetectionADBench High Dimensionality d > 100 (15 datasets)
ROC-AUC73.98
10
Semi-supervised Anomaly DetectionADBench 100% labeled anomaly ratio
Mean Rank (Rk)2.2
9
Semi-supervised Anomaly DetectionADBench 25% labeled anomaly ratio
Mean Rank (Rk)2.5
9
Semi-supervised Anomaly DetectionADBench 5% labeled anomaly ratio
Mean Rank2.9
9
Semi-supervised Anomaly DetectionADBench 1% labeled anomaly ratio
Mean Rank2.7
9
Semi-supervised Novelty DetectionADBench Novelty
Mean Rank2.6
9
Anomaly DetectionADBench One-Class
Mean Rank3.3
9
Anomaly DetectionADBench Unsupervised
Mean Rank3.9
9
Anomaly DetectionADBench annthyroid
Average AUC PR82.88
9
Showing 25 of 28 rows