NEXT-GEN RANSOMWARE DEFENSE: MACHINE LEARNING- POWERED THREAT DETECTION

Abstract

Ransomware attacks pose a significant and escalating threat to individuals and organizations, causing substantial financial and operational disruptions. This research explores the application of machine learning techniques for the proactive prediction of ransomware activity. By analyzing a comprehensive dataset of system behaviors, network traffic, and file system modifications, we develop  predictive  models capable of identifying potential ransom ware attacks before encryption occurs. We employ a range of machine learning algorithms, including Random Forest,  Lazy Predict, to classify malicious activity. Our methodology incorporates feature engineering to extract relevant indicators of ransom ware behavior, and we evaluate the performance of our models using rigorous testing and validation of datasets using LIME. In order to improve the detection rate of ransomware, the data imbalance was addressed by utilizing the SMOTE- Tomek method, which allowed for a more robust machine learning prediction model. The results demonstrate the effectiveness of machine learning in enhancing ransomware detection and mitigation, offering a valuable tool for strengthening cyber security defenses.