IntelliGuard: Intelligence on Cyber Risk for Commercial Protection

Abstract

The Internet of Things paradigm has transformed connectivity, but it has also introduced previously unheard-of security risks. The classification of assaults using a unfathomable erudition approach and other machine learning approaches is the prime focal point of this study. Using machine learning (ML) technique such a shaphazard Forest (RF),conclusion tree(DT), Extra Tree Trainer (ETC), Support Vector Machine(SVM),and k-Nearest Neighbor (KNN) and Deep Learning (DL) design concepts, including Deep Neural Network (DNN), the study analyzes various attack types in IoT environments and proposes a binary and multiclass classification framework. Benchmark datasets with real-world IoT attack scenarios, such as Edge-IIoT set, are used for experiments. The dataset is preprocessed using Principal Componenet Analysis (PCA) for choosing features, Synthetic Minority Over sampling Technique to solve class imbalance, and Standard Scaling for feature scaling. These approaches' comparative performance and efficacy are examined. The outcomes demonstrate how well machine learning strategies generalize across various attack classes and how well the DL model manages intricate assault patterns. The DNN model yields the most excellent results with 100% accuracy for binary classification, 96.15% accuracy for 6-class classification, and 94.68% effectiveness for 15-class classification. A ten-fold cross validation approach has also been used to check for over fitting in the model. This work helps with the improvement of security mechanisms by offering insights into the selection of appropriate strategies for the binary and multiclass breakdown of threats.