Evolution of intrusion detection using machine learning and deep learning

Intrusion Detection Scheme (IDS) is one of the counter measures against harmful attacks. The rapid development of technology not only make life easier, but it also raises a number of security concerns. Despite decades of research, offered IDSs still have problems increasing detection accuracy, reducing false alarm rates, and identifying new threats. Many scholars have determined on building DSs that use machine learning methods to cope with the difficulties mentioned above. Machine learning technique can immediately track the key differences between regular and irregular data. Machine Learning (ML) approach have recently been implemented in DS to identify and classify probable problems. This article discusses ML techniques used in IDS for a range of applications, including fog computing, the Internet of Things (IoT), big data, smart cities, and 5G networks. In addition, this research tries to classify intrusions using machine learning techniques such as Random Forest, Linear Discriminant Analysis (LDA), and Classification and Regression Trees (CART). The most frequently used machine learning algorithms in IDSs, metrics, and benchmark datasets are reviewed, followed by a demonstration of how to apply machine learning and deep learning techniques to address major IDS issues, utilizing sample literature as a roadmap.