Gradient Boosting Feature Selection With Machine Learning Classifiers for Intrusion Detection on Power Grids

Smart grids rely on SCADA (Supervisory Control and Data Acquisition) systems to monitor control complex electrical networks in order provide reliable energy homes industries. However, the increased inter-connectivity remote accessibility of expose them cyber attacks. As a consequence, developing effective security mechanisms is priority protect network from internal external We propose an integrated framework for Intrusion Detection System (IDS) smart which combines feature engineering-based preprocessing with machine learning classifiers. Whilst most techniques fine-tune hyper-parameters improve detection rate, our approach focuses selecting promising features dataset using Gradient Boosting Feature Selection (GBFS) before applying classification algorithm, combination improves not only rate but also execution speed. GBFS uses Weighted Importance (WFI) extraction technique reduce complexity implement evaluate various decision-tree based after obtaining power grid through module, show that this optimizes False Positive Rate (FPR) time.