Anomaly Detection on WBC
99.5ROCAUCBest Unsupervised
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Best Unsupervisedsource=[9]2025.12 | 99.5 | 92.3 | — | |
| PaLD2025.12 | 94.8 | 48 | — | |
| CTADBase Detector=DRL2026.02 | 1 | — | — | |
| DRL2026.02 | 1 | — | — | |
| CTADBase Detector=DRL2026.02 | 1 | — | — | |
| knn2025.12 | 1 | 1 | — | |
| PCA2026.02 | 0.9989 | — | — | |
| IForest2026.02 | 0.9983 | — | — | |
| IForest2026.02 | 0.998 | — | — | |
| ECOD2026.02 | 0.998 | — | — | |
| CTADBase Detector=KNN2026.02 | 0.9978 | — | — | |
| AutoEncoder2026.02 | 0.9976 | — | — | |
| DRL2025.10 | 0.9963 | — | — | |
| SLAD2026.02 | 0.996 | — | — | |
| DRL2026.02 | 0.996 | — | — | |
| GOAD2026.02 | 0.995 | — | — | |
| PCA2026.02 | 0.994 | — | — | |
| ICAD-LLMAD Method Category=Task-Specific, Backbone=Qwen2.5-0.5B, Reference Set Size (K)=52025.12 | 0.9906 | — | — | |
| ICAD-LLMMethod Category=Task-Specific2025.12 | 0.9906 | — | — | |
| ICL2026.02 | 0.9898 | — | — | |
| KNN2026.02 | 0.989 | — | — | |
| LLM-DASMethod Variant=DRL2025.10 | 0.9883 | — | — | |
| DeepSVDD2026.02 | 0.9879 | — | — | |
| CTADBase Detector=MCM2026.02 | 0.9863 | — | — | |
| DTE2026.02 | 0.9852 | — | — | |
| RCA2026.02 | 0.983 | — | — | |
| REPEN2026.02 | 0.983 | — | — | |
| MCM2026.02 | 0.9814 | — | — | |
| MCM2025.10 | 0.9814 | — | — | |
| ICAD-LLMAD Method Category=Universal, Backbone=Qwen2.5-0.5B, Reference Set Size (K)=52025.12 | 0.9794 | — | — | |
| ICAD-LLMMethod Category=Universal2025.12 | 0.9794 | — | — | |
| ECOD2026.02 | 0.9793 | — | — | |
| MCMAD Method Category=Task-Specific2025.12 | 0.9789 | — | — | |
| MCMMethod Category=Task-Specific2025.12 | 0.9789 | — | — | |
| PLAGLLM=Qwen3-14B2026.04 | 0.9779 | 0.9019 | 82.75 | |
| ICAD-LLMModality=Tabular, Generalization (Unseen Datasets)=true2025.12 | 0.976 | — | — | |
| AutoEncoder2026.02 | 0.9737 | — | — | |
| AutoEncoder2025.10 | 0.9737 | — | — | |
| KNN2026.02 | 0.9728 | — | — | |
| RTTAD2026.05 | 0.9723 | — | — | |
| IForest2026.02 | 0.9715 | — | — | |
| IForest2025.10 | 0.9715 | — | — | |
| PLAGLLM=GPT-3.5-turbo2026.04 | 0.971 | 0.8649 | 75.5 | |
| CausalTAD-135Mbackbone=SmolLM-135M2026.02 | 0.969 | — | — | |
| LOF2025.10 | 0.967 | — | — | |
| PLAGLLM=DeepSeek-V2.52026.04 | 0.9669 | 0.8639 | 77.9 | |
| OCSVM2026.02 | 0.9667 | — | — | |
| PCA2026.02 | 0.9667 | — | — | |
| OCSVM2025.10 | 0.9667 | — | — | |
| PCA2025.10 | 0.9667 | — | — | |
| AnoLLM-135Mbackbone=SmolLM-135M2026.02 | 0.966 | — | — | |
| MCM2026.05 | 0.9643 | — | — | |
| AnoLLMModel Size=135M2025.10 | 0.964 | — | — | |
| Gaussian noise2026.04 | 0.9639 | 0.8224 | 77.87 | |
| OCSVM2026.02 | 0.9637 | — | — | |
| DeepSVDD2026.02 | 0.9633 | — | — | |
| DeepSVDD2025.10 | 0.9633 | — | — | |
| PLAGLLM=GPT-42026.04 | 0.9622 | 0.8817 | 78.26 | |
| NPT-AD2025.10 | 0.9619 | — | — | |
| CTADBase Detector=KNN2026.02 | 0.9606 | — | — | |
| ICL2026.02 | 0.957 | — | — | |
| Two-Stage LKPLO (SVM-like)validation=5-fold cross-validation2025.10 | 0.957 | — | — | |
| NeuTraLAD2026.02 | 0.9568 | — | — | |
| KNN2026.02 | 0.9536 | — | — | |
| KNN2025.10 | 0.9536 | — | — | |
| AnoLLMModel Size=360M2025.10 | 0.952 | — | — | |
| CBLOFvalidation=5-fold cross-validation2025.10 | 0.951 | — | — | |
| SOEL2026.04 | 0.9499 | 0.8338 | 71.42 | |
| CTADBase Detector=MCM2026.02 | 0.9466 | — | — | |
| TabSyn2026.04 | 0.9463 | 0.8939 | 80 | |
| MCM2026.02 | 0.9419 | — | — | |
| TabDDPM2026.04 | 0.9402 | 0.8791 | 79.06 | |
| OCSVMvalidation=5-fold cross-validation2025.10 | 0.94 | — | — | |
| KDEvalidation=5-fold cross-validation2025.10 | 0.939 | — | — | |
| KNNvalidation=5-fold cross-validation2025.10 | 0.937 | — | — | |
| Two-Stage LKPLO (Robust-Z)validation=5-fold cross-validation2025.10 | 0.937 | — | — | |
| LLM-DASMethod Variant=PCA2025.10 | 0.9333 | — | — | |
| KPCAvalidation=5-fold cross-validation2025.10 | 0.932 | — | — | |
| WRPOvalidation=5-fold cross-validation2025.10 | 0.931 | — | — | |
| DeepSVDD2026.02 | 0.93 | — | — | |
| D2H-ADD=10,000, K=322026.06 | 0.928 | — | 90.3 | |
| KRPDvalidation=5-fold cross-validation2025.10 | 0.928 | — | — | |
| IForestvalidation=5-fold cross-validation2025.10 | 0.927 | — | — | |
| KODvalidation=5-fold cross-validation2025.10 | 0.927 | — | — | |
| LOFvalidation=5-fold cross-validation2025.10 | 0.924 | — | — | |
| ICL2026.05 | 0.9234 | — | — | |
| RPDvalidation=5-fold cross-validation2025.10 | 0.919 | — | — | |
| ODHDD=10,000, K=322026.06 | 0.916 | — | 84 | |
| NeuTraLAD2026.02 | 0.9133 | — | — | |
| FITYMITraining Regime=Pre-trained, Backbone=ViT-B_162022.05 | 0.912 | — | — | |
| ICL2026.02 | 0.908 | — | — | |
| ICL2025.10 | 0.908 | — | — | |
| GReaT2026.04 | 0.908 | 0.8133 | 74.99 | |
| GOAD2026.05 | 0.9032 | — | — | |
| NNG-Mix2026.04 | 0.8975 | 0.4689 | 38.71 | |
| TTAD2026.05 | 0.894 | — | — | |
| DRL2026.05 | 0.8887 | — | — | |
| DIF2026.05 | 0.8877 | — | — | |
| ECOD2026.02 | 0.8747 | — | — | |
| ECOD2025.10 | 0.8747 | — | — |