An Intelligent Approach for Intrusion Detection using Snake Optimizer and Random Forest
Salahaldeen Duraibi · 2022
In recent years, many researchers have used various Machine Learning (ML) models that have demonstrated the power of such methods on Intrusion Detection (ID) and thus helped classifying network packets as normal system behavior or an attack. This paper presents a novel SO-RF model that combines Snake Optimizer (SO) and Random Forest (RF) for ID. The SO Meta-Heuristics (MH) algorithm is employed to select Optimal Feature Subset (OFS) from large datasets and the resulted OFS is used by the RF model to improve learning process and classification accuracy. The SO-RF is validated on two datasets for ID: KDD CUP99 and NSL-KDD. Results show that the introduced SO-RF achieves better performance outcomes compared to the RF, SVM, and several other reported models in the literature for ID.