A Consolidated Decision Tree-Based Intrusion Detection System for Binary and Multiclass Imbalanced Datasets

The widespread acceptance and increase of the Internet mobile technologies have revolutionized our existence. On other hand, world is witnessing suffering due to technologically aided crime methods. These threats, including but not limited hacking intrusions are main concern for security experts. Nevertheless, challenges facing effective intrusion detection methods continue closely associated with researcher’s interests. This paper’s contribution present a host-based system using C4.5-based detector on top popular Consolidated Tree Construction (CTC) algorithm, which works efficiently in presence class-imbalanced data. An improved version random sampling mechanism called Supervised Relative Random Sampling (SRRS) has been proposed generate balanced sample from high-class imbalanced dataset at detector’s pre-processing stage. Moreover, an multi-class feature selection designed developed as filter component IDS datasets’ ideal outstanding features efficient detection. validated state-of-the-art systems. results show accuracy 99.96% 99.95%, considering NSL-KDD CICIDS2017 34 features.