A Hybrid Intrusion Detection with Decision Tree for Feature Selection

Due to the size and nature of intrusion detection datasets, systems (IDS) typically take high computational complexity examine features data identify intrusive patterns. Data preprocessing techniques such as feature selection can be used reduce by eliminating irrelevant redundant in dataset. The objective this study is analyze efficiency effectiveness some approaches namely, wrapper-based filter-based modeling approaches. To achieve that, a hybrid algorithm combination with wrapper filter processes designed. We propose decision tree guide process. Five machine learning algorithms are on methods build IDS models using UNSW-NB15 three information gain, gain ratio, relief for comparison determine proposed approach. Furthermore, fair other state-of-the-art also performed. experimental results show that our approach quite effective works, however, it takes time whilst achieves similar results. Our work revealed unobserved issues about conformity