DETECTING AND CLASSIFYING RANSOMWARE WITH MACHINE LEARNING ALGORITHMS

Abstract

Cybersecurity is always at risk from malicious assaults, malware, and ransomware families, which may wreak havoc on several industries' worth of computer networks, data centers, websites, and mobile apps. Conventional anti-ransomware technologies are rendered ineffective by more complex ransomware assaults. Therefore, cutting-edge methods, including traditional and neural network-based designs, may be invaluable to developing new ransomware remedies. In this paper, we provide a framework for detecting and preventing ransomware that uses a number of machine learning methods, such as designs based on neural networks and feature selection, to categorize security levels. A few of ransomware characteristics were subjected to several machine learning methods, including Decision Tree (DT), Random Forest (RF), Naïve Bayes (NB), Logistic Regression (LR), and NN-based classifiers. For the purpose of thoroughly evaluating our suggested technique, we used a single ransomware dataset. Experimental data show that RF classifiers beat competing approaches on accuracy, F-beta, and precision measures. Information security, AI, NN, ransomware categorization