Table 3.
Comparative results of GAN-based and MVE-based models using Sensor Groups 1 and 2. Higher performance indicator values are marked in with ∗.
Sensors | Amplitude | Models | False alarm | Event detection rate | F1 score | Average detection time (min) |
---|---|---|---|---|---|---|
Group 1 | 1.0–1.5 | GAN | 2∗ | 0.86∗ | 0.42∗ | 53.8∗ |
MVE | 2∗ | 0.57 | 0.36 | 54.5 | ||
1.5–2.0 | GAN | 3 | 0.86∗ | 0.53∗ | 47.3 | |
MVE | 2∗ | 0.71 | 0.41 | 47.0∗ | ||
2.0–2.5 | GAN | 3 | 0.86∗ | 0.56∗ | 40.0∗ | |
MVE | 2∗ | 0.86∗ | 0.54 | 44.2 | ||
2.5–3.0 | GAN | 3 | 0.86 | 0.59 | 40.0 | |
MVE | 2∗ | 1.00∗ | 0.66∗ | 37.3∗ | ||
Group 2 | 1.0–1.5 | GAN | 1∗ | 1.00∗ | 0.34∗ | 84.9 |
MVE | 15 | 0.43 | 0.22 | 74.0∗ | ||
1.5–2.0 | GAN | 2∗ | 0.86∗ | 0.41∗ | 71.7∗ | |
MVE | 16 | 0.71 | 0.31 | 78.8 | ||
2.0–2.5 | GAN | 3∗ | 1.00∗ | 0.44∗ | 69.0∗ | |
MVE | 15 | 0.86 | 0.38 | 73.7 | ||
2.5–3.0 | GAN | 2∗ | 1.00∗ | 0.45∗ | 67.6∗ | |
MVE | 16 | 0.86 | 0.42 | 67.8 |