Malware Detection in PDF Files Using Machine Learning

The proliferation of Portable Document Format (PDF) files as a standard for document exchange has made them a prime target for cybercriminals who embed malicious code to exploit system vulnerabilities. Traditional malware detection techniques, such as signature-based and heuristic-based methods, struggle to detect advanced and zero-day threats due to evolving obfuscation and encryption techniques. This research presents a machine learning-based approach to detecting malware in PDF files using feature extraction techniques. By analyzing key attributes such as embedded JavaScript, file structure anomalies, and metadata inconsistencies, the proposed system classifies PDFs as either benign or malicious. The study demonstrates the effectiveness of a Random Forest classifier in distinguishing between clean and infected files, achieving high accuracy while minimizing false positives. This research contributes to enhancing cyber security defenses by integrating data-driven methodologies to improve threat detection mechanisms for PDF-based malware