A Scalable, Ensemble Approach for Building and Visualizing Deep Code-Sharing Networks

Explore a cutting-edge approach to analyzing malware code-sharing relationships in this 22-minute Black Hat conference talk. Discover how to leverage an obfuscation-resilient ensemble similarity analysis method that scales to millions of samples and withstands various obfuscation techniques. Learn about the algorithm's development, evaluation process, and its superior performance against competing malware cluster recognition techniques. Gain insights into implementing the algorithm and visualizing large-scale malware networks. Get introduced to a Python machine learning library for detecting feature frequencies across billions of items on standard hardware, which will be released with the conference materials.