Machine Learning for Detecting Malware in PE Files

TL;DR

This paper reviews and evaluates machine learning techniques for detecting malware in PE files, using a large benchmark dataset to assess feature effectiveness and detection performance.

Contribution

It provides a comprehensive evaluation of machine learning methods and features for PE malware detection on a large benchmark dataset.

Findings

Certain features significantly improve detection accuracy

Machine learning models outperform traditional signature-based methods

Evaluation highlights strengths and limitations of current approaches

Abstract

The increasing number of sophisticated malware poses a major cybersecurity threat. Portable executable (PE) files are a common vector for such malware. In this work we review and evaluate machine learning-based PE malware detection techniques. Using a large benchmark dataset, we evaluate features of PE files using the most common machine learning techniques to detect malware.