Anomaly Detection on Synthetic Manufacturing Dataset (test)
94.4AccuracyRandom Forest
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Random Forest2025.01 | 94.4 | 94.3 | 100 | 97.1 | 98.2 | 99.9 | |
| Vanilla LSTM2025.01 | 94.3 | 94.3 | 99.9 | 97 | 91.3 | 99.4 | |
| CNN-LSTMvariant=Ablation: base CNN-LSTM2025.01 | 93.9 | 94 | 99.8 | 96.8 | 83 | 98.1 | |
| ANSR-DT (Full)variant=Full model2025.01 | 93.5 | 93.6 | 99.9 | 96.6 | 95.5 | 99.6 | |
| CNN-LSTM+Attnvariant=Ablation: CNN-LSTM with Attention2025.01 | 93.5 | 93.6 | 99.9 | 96.6 | 84.4 | 97.6 | |
| CNN-LSTM+Symbolicvariant=Ablation: CNN-LSTM with Symbolic logic2025.01 | 93.5 | 93.6 | 99.9 | 96.6 | 95.5 | 99.6 | |
| Transformer2025.01 | 93.4 | 93.4 | 100 | 96.6 | 95.8 | 99.7 | |
| Pure Symbolic2025.01 | 71.7 | 99.7 | 70 | 82.2 | 83.5 | 98.9 | |
| One-Class SVM2025.01 | 31.7 | 100 | 26.8 | 42.3 | 40 | 93.9 | |
| LSTM Autoencoder2025.01 | 25.8 | 100 | 20.5 | 34.1 | 65.1 | 97.2 | |
| Isolation Forest2025.01 | 23.8 | 90.6 | 20.6 | 33.6 | 27.8 | 91.2 | |
| LOF2025.01 | 10.2 | 100 | 3.9 | 7.4 | 34.5 | 92.6 |