Table 8.
Linear Support Vector Machine (SVM) performance.
| Linear SVM | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Dataset | Score | Desktop | MCUs | |||||||
| Type | Name | Python | C | F0 | F3 | F4 | F7 | H7 | L4 | |
| Binary | Heart | Acc. | 84% | 84% | 84% | 84% | 84% | 84% | 84% | 84% |
| Inf. Time | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||
| Virus | Acc. | 94% | 94% | 94% | 94% | 94% | 94% | 94% | 94% | |
| Inf. Time | <1 ms | <1 ms | 1 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||
| Sonar | Acc. | 78% | 78% | 78% | 78% | 78% | 78% | 78% | 78% | |
| Inf. Time | <1 ms | <1 ms | 3 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||
| Multiclass | Peugeot_Target 14 | Acc. | 91% | 91% | 91% | 91% | 91% | 91% | 91% | 91% |
| Inf. Time | <1 ms | <1 ms | 2 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||
| Peugeot_Target 15 | Acc. | 90% | 90% | 90% | 90% | 90% | 90% | 90% | 90% | |
| Inf. Time | <1 ms | <1 ms | 2 ms | <1 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||
| Regress | EnviroCar | R2 | 0.99 | 0.99 | 0.99 | 0.99 | 0.99 | 0.99 | 0.99 | 0.99 |
| Inf. Time | <1 ms | <1 ms | 5 ms | 3 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||
| AQI | R2 | 0.73 | 0.73 | 0.73 | 0.73 | 0.73 | 0.73 | 0.73 | 0.73 | |
| Inf. Time | <1 ms | <1 ms | 5 ms | 3 ms | <1 ms | <1 ms | <1 ms | <1 ms | ||