Cybersecurity Detection Model using Machine Learning Techniques

Abstract

The use of machine learning methods in cybersecurity is only one of many examples of how this once-emerging innovation has entered the mainstream. Anomaly-based identification of common assaults on vital infrastructures is only one instance of the various applications of malware analysis. Scholars are using machine learning-based identification in numerous cybersecurity solutions since signature-based approaches are inadequate at identifying zero-day threats or even modest modifications of established assaults. In this work, we introduce the machine-learning models-based security framework to detect cyber-attacks. This paper used three machine learning models Logistic Regression, Random Forest, and K-Nearest Neighbor This framework not only reduces the computational difficulty of the framework by minimizing the feature parameters, but it also performs well in terms of accuracy in forecasting unknown scenarios in the tests. Finally, we ran trials using cybersecurity datasets to measure the machine learning model's performance using metrics including precision, recall, and accuracy.