Published Paper Details:

This systematic review examines the application of Artificial Intelligence (AI) and Machine Learning (ML) techniques in Intrusion Detection Systems (IDS). The review analyses 162 peer-reviewed studies published between 2015 and 2024, focusing on supervised learning, unsupervised learning, and deep learning approaches used in Network-based IDS (NIDS), Host-based IDS (HIDS), and Hybrid IDS architectures. The findings indicate that deep learning models, especially Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), and Generative Adversarial Networks (GAN), outperform traditional ML methods on benchmark datasets such as NSL-KDD, UNSW-NB15, and CICIDS2017, achieving F1-scores above 0.98. However, challenges such as class imbalance, adversarial attacks, concept drift, high false positive rates, and limited dataset generalization continue to affect practical deployment. Emerging technologies including Federated Learning, Explainable AI (XAI), and Graph Neural Networks (GNN) show promising results for future IDS development. This review provides a structured overview of AI-based IDS techniques, benchmark datasets, current challenges, and future research directions.