Machine Learning-Based Detection for Unauthorized Access to IoT Devices

The Internet of Things (IoT) has become widely adopted in businesses, organizations, and daily lives. They are usually characterized by transferring processing sensitive data. Attackers have exploited this prospect IoT devices to compromise user data’s integrity confidentiality. Considering the dynamic nature attacks, artificial intelligence (AI)-based techniques incorporating machine learning (ML) promising for identifying such attacks. However, dataset being utilized features engineering techniques, kind classifiers play significant roles how accurate AI-based predictions are. Therefore, environment, there is a need contribute more context evaluating different on datasets that effectively capture environment’s properties. In paper, we evaluated various ML models with consideration both binary multiclass classification validated new dedicated dataset. Moreover, investigated impact including correlation analysis information gain. experimental work conducted bagging, k-nearest neighbor (KNN), J48, random forest (RF), logistic regression (LR), multi-layer perceptron (MLP) revealed RF achieved highest performance across all experiment sets, receiver operating characteristic (ROC) 99.9%.