A Comprehensive Review of Machine Learning and Deep Learning Techniques for Addressing Class Imbalance Issues in Network Intrusion Detection Systems

Nowadays, a lot of information is shared between various connecting devices due to the quick development of internet technologies and the growing reliance on online services. Security systems are one of the most important scientific topics in today’s computer network, so data should be transmitted safely between connected devices. The intrusion detection system is crucial for network security because it can identify and stop malicious activities. IDS still has trouble identifying novel intrusions due to data imbalance problems and a high false alarm rate. Many different methods have been applied to either prevent or detect intrusions. We discussed several research problems as well as potential future directions for NIDS. This study investigates ML-based and DL-based intrusion detection strategies and dataset selection.