gholamreza Ahmadi

As cyberattacks such as distributed denial of service (DDoS) attacks, intrusions, and malware become more complex and diverse, the need for advanced methods to detect these threats in computer networks has become more apparent than ever. Machine learning has become a key tool in cybersecurity due to its ability to analyze complex patterns and identify anomalies [1]. This paper reviews the application of machine learning models—including supervised learning (e.g., SVM), deep learning (e.g., LSTM), and hybrid models—in cyberattack detection. The proposed models are evaluated using standard NSL-KDD and CICIDS2017 datasets, and the results show that the hybrid models achieve 94% accuracy for detecting attacks in IoT networks [10]. Challenges such as adversarial attacks against machine learning models and issues related to unbalanced data are also discussed [5]. Finally, solutions for improving detection systems and future research directions are presented.