SMOTE Integrated Adaptive Boosting Framework for Network Intrusion Detection

Network abnormalities may occur due to enormous reasons, such as user irregular behavioral patterns, network system failure, attacker malicious activities, botnets, or malicious software. The importance of information management and data processing systems has changed the enormous volume of data and its incremental increase. An IDS monitors and examines data to detect unauthorized entries into a system or network. In this article, the Ada-Boost ensemble learning technique is proposed with SMOTE to identify the anomalies in the network. The Ada-Boost algorithm is utilized mainly in the classification task, and SMOTE handles the class imbalance problem. The suggested approach outperformed various ML algorithms and ensemble learning approaches in relation of precision, recall, F1-score, and accuracy with 0.999 and 99.97% respectively when investigated with the NSL-KDD dataset.