Network Intrusion Detection with Two-Phased Hybrid Ensemble Learning and Automatic Feature Selection

The use of network connected devices has grown exponentially in recent years revolutionizing our daily lives. However, it also attracted the attention cybercriminals making attacks targeted towards these increase not only numbers but sophistication. To detect such attacks, a Network Intrusion Detection System (NIDS) become vital component applications. produce large scale high-dimensional data which makes difficult to accurately various known and unknown attacks. Moreover, complex nature feature selection process NIDS challenging task. In this study, we propose machine learning based with Two-phased Hybrid Ensemble Automatic Feature Selection. proposed framework leverages four different classifiers perform automatic on their ability most significant features. two-phased hybrid ensemble algorithm consists two phases, first phase constructed using built from an adaptation One-vs-One framework, second combinations attack classes. was evaluated well-referenced datasets for both wired wireless applications, results demonstrate that combined engine superior detection capability compared other similar studies found literature.