Hybrid Optimization Based Convolutional Neural Network for Intrusion Detection System

Abstract

Intrusion Detection System (IDS) is one of the common deep learning (DL) techniques that are used to find and identify outliers to prevent adversarial attacks, fraud, and network intrusions. This paper proposed a hybrid Particle Swarm Optimization and Grey Wolf Optimization (HPSOGWO) based Convolutional Neural Network (CNN) model for intrusion detection systems. The proposed HPSOGWO is evaluated on the NSL-KDD dataset which contains 5 different classes and 41 features. The PSO is good for global exploration and GWO is for local exploitation. The hybrid PSO and GWO algorithm achieves a better balance between exploration and exploitation and enhances convergence speed. The CNN is utilized to enhance the system's capability to identify and classify intrusion accurately and effectively. The proposed HPSOGWO-based CNN model attains better results by utilizing evaluation metrics like accuracy, precision, recall, specificity, and f1-score values about 0.9918, 0.9852, 0.9908, 0.9879 and 0.9767 correspondingly which is comparatively higher than existing techniques like Chicken Swarm Optimization based Deep Long Short-Term Memory (ChCSO based Deep LSTM), Deep Neural Network (DNN) and LSTM.