Machine Learning-Based Intrusion Detection for Zero-Day Ransomware in Unseen Data

Ransomware continues to pose a significant threat to global digital infrastructure, as its complexity and capacity for evasion have advanced rapidly in recent years. The novel approach proposed in this research integrates machine learning into intrusion detection systems, offering a robust solution to identify zero-day ransomware attacks in unseen data. Through rigorous experimentation with multiple machine learning algorithms, including Random Forest, Support Vector Machines, and Neural Networks, the study demonstrates that machine learning models can effectively detect previously unknown ransomware behavior through the analysis of system anomalies. Performance metrics such as accuracy, precision, and recall illustrate the strengths and trade-offs associated with different models, while results from zero-day simulations highlight the adaptability of machine learning techniques in handling novel threats. The findings demonstrate the potential for machine learning to enhance cybersecurity defenses by providing more dynamic and scalable detection methods for emerging ransomware attacks.