Table 7.
Experimental results for Testing_2_sampled dataset altered with peak anomalies. In bold and italics are reported the best results for the Sparse U-Net and the baseline, respectively.
| Neural Model | Threshold | Accuracy | Precision | Recall | F-Measure |
|---|---|---|---|---|---|
| Deep Autoencoder ( baseline ) | 98th percentile | 0.852 | 0.395 | 0.973 | 0.562 |
| max value | 0.945 | 0.986 | 0.449 | 0.617 | |
| max + tolerance | 0.927 | 1.000 | 0.249 | 0.399 | |
| Sparse U-Net ( Proposed Model ) | 98th percentile | 0.875 | 0.438 | 0.966 | 0.603 |
| max value | 0.968 | 0.952 | 0.711 | 0.814 | |
| max + tolerance | 0.961 | 1.000 | 0.604 | 0.753 |