Machine Learning Techniques for Enhanced IOT Security

The rapid growth of internet of things (IoT) made data security and privacy the most challenging tasks. Search technology was introduced to access large amounts of data; later, access mechanisms based on attributes, policies, and authorization were evolved. However, these were insufficient to provide user data security, causing unauthorized access and cyber-attacks. Introducing machine learning (ML) techniques provides enhanced security against various types of cyber attacks, such as normal attacks, DDOS, man-in-middle-man attacks, keylogging, data theft, and SSR. This paper presents a review and analysis of state-of-the-art machine-learning techniques used to enhance the performance of various methods required to detect threats and cyber attacks in IoT networks effectively. Popular machine learning techniques, correlation accuracy metric (CorrACC), and correlation area under the curve (CorrAUC) for building an effective feature set are presented, and a detailed discussion on dataset validation and performance evaluation on precision, accuracy, sensitivity, and specificity by applying ML techniques is included.