Author : Mani Gopalsamy

Cybersecurity is rapidly embracing ML. Integrating ML into cybersecurity mainly aims to improve the effectiveness, scalability, and actionability of malware detection compared to more conventional approaches that depend on human intervention. Problems with ML need well-managed theoretical and methodical approaches in the cybersecurity sector. The increasing prevalence of cyber threats necessitates effective strategies for malware detection within cybersecurity frameworks. Using the EMBER v2017 dataset—this study intends to develop and assess ML methods for malware attack detection and classification. This research used machine learning classification algorithms Neural Network (NN), Random Forest (RF), and SVM (Support vector machine) and evaluated the performance of these models in terms of F1 score, precision, accuracy, and recall. The Neural Network model exceeds the others, with an accuracy of 97.53% and a precision of 98.85%, whereas RF has a lesser accuracy of 84.3%. These findings underscore the importance of using powerful machine-learning techniques to improve cybersecurity safeguards against emerging threats. The work contributes to the field by providing a detailed examination of the performance of several malware detection techniques, as well as recommendations for future research and practical cybersecurity applications.

Keywords: Malware Detection, Cybersecurity, EMBER Dataset, machine learning, classification algorithms.