Guarding the Digital Frontier: A Logistic Regression Approach to Malware Detection

Taking new computing resources, a constantly evolving field, the malware threat is so widespread that a strong detection requires a creative response. The traditional methods of malware detection that include signature or shape-based methods do not perform well against different types of malware developed. So, we propose a malware detection model consisting of two parts: one for detecting malware based on PE (Portable Executable) header features, and another for detecting malicious websites based on URL features. The model uses random forest classifier for malware detection based on PE header features, statistical method called logistic regression for malicious website detection based on URL features thereby aiming to provide a fast, reliable, and trustworthy solution for guarding the digital frontier.