Model Machine Learning SVM (Support Vector Machine) untuk Deteksi Anomali pada Sistem Kelistrikan Perusahaan Kerajinan Kayu GS4

Anomaly detection in electrical systems is crucial to prevent operational disruptions and equipment damage, especially in small industries such as handicraft companies. This study aims to develop an electrical anomaly detection model using Support Vector Machine (SVM) based on current, voltage, and temperature parameters. Data were collected in real-time using sensors installed at strategic points in the company's electrical network. Anomaly criteria were determined based on normal operating limits: current (8.2–10 A), voltage (198–242 V), and temperature (30–70°C). The SVM model was trained using a dataset classified into normal and anomalous conditions. Model evaluation was conducted using accuracy, precision, recall, and F1-score metrics to assess the anomaly detection performance. Model evaluation was performed using accuracy, precision, recall, and F1-score metrics to assess the accuracy of anomaly detection. The results showed that the SVM model was able to identify anomalies with high accuracy, namely with an Accuracy value of 96.5%. Precision of 94.8% and Recall of 92.3%.