Table 1 Confusion matrix for testing dataset for different SOC cut-off limit.
From: Internal short circuit detection in Li-ion batteries using supervised machine learning
| Â | Â | Â | NiNÂ =Â 100% | Â | Â | Â | NiNÂ =Â 97.97% |
SOCÂ =Â 5% | Faulty | Healthy | FiFÂ =Â 98.45% | SOCÂ =Â 15% | Faulty | Healthy | FiFÂ =Â 97.67% |
Faulty | TN = 127 | FP = 2 | False Alarm = 0.0% | Faulty | TN = 126 | FP = 3 | False Alarm = 2.03% |
Healthy | FN = 0 | TP = 148 | Miss-detection = 1.55 | Healthy | FN = 3 | TP = 145 | Miss-detection = 2.33% |
| Â | Â | Â | NiNÂ =Â 97.97% | Â | Â | Â | NiNÂ =Â 93.92% |
SOCÂ =Â 30% | Faulty | Healthy | FiFÂ =Â 97.67% | SOCÂ =Â 50% | Faulty | Healthy | FiFÂ =Â 97.67% |
Faulty | TN = 126 | FP = 3 | False Alarm = 2.03% | Faulty | TN = 126 | FP = 3 | False Alarm = 6.08% |
Healthy | FN = 3 | TP = 145 | Miss-detection = 2.33 | Healthy | FN = 9 | TP = 139 | Miss-detection = 2.33% |
