Review on Intrusion Detection System Using Deep Learning and Machine Learning

With the increasing complexity and sophistication of cyber-attacks, intrusion detection systems (IDSs) play a key role in safeguarding computer networks from unauthorized access and malicious activities. Traditional IDSs have limitations in accurately identifying evolving attack patterns due to their static rule-based nature. To overcome these limitations, we explored the extensive approach of machine learning (ML) and deep learning (DL) techniques into IDS. This paper presents a review on strengths and weaknesses of intrusion detection using deep learning and machine learning techniques and different datasets. such as the need for large-scale labelled datasets, interpretability of deep learning models, and the impact of adversarial attacks on DDos. The review addresses the challenges and further improves the accuracy and robustness of IDSs against emerging cyber threats using hybrid approaches that combine ML and DL techniques for intrusion detection. Several integration strategies, such as feature extraction, model training and real time datasets, are explored, along with their respective benefits and challenges.