Cyber Attack Prediction Framework Transitioning from Traditional Machine Learning to Generative AI Models
DOI:
https://doi.org/10.46647/rdems0205001Keywords:
Cyber Security, Cyber Attack Prediction, Machine Learning, Generative AI, Deep Learning, Network Security, Intrusion Detection System, Data Preprocessing, Feature Extraction, Random Forest, Support Vector Machine, GAN, Autoencoder, Anomaly Detection, Zero-Day Attacks, Threat Detection, Data Analysis, Model Training, Classification, Network Traffic AnalysisAbstract
Cyber attacks are increasing rapidly due to the expansion of digital infrastructure, cloud computing, and connected devices. Traditional machine learning techniques have been widely used to detect cyber threats; however, they often struggle with evolving attack patterns and zero-day vulnerabilities. This research proposes a cyber attack prediction framework that transitions from conventional machine learning approaches to Generative Artificial Intelligence models for improved threat detection and prediction. The system integrates network traffic analysis, behavioral monitoring, and deep generative models to learn complex patterns from cybersecurity datasets. Generative AI models can simulate potential attack scenarios and identify hidden threats before they occur. The proposed framework enhances prediction accuracy, adaptability, and automated threat intelligence, providing a proactive cybersecurity solution capable of identifying sophisticated cyber attacks in modern digital environments.