GTC is all about four days of discovery. Come explore what's driving transformation in your industry—from the power of AI to the collaborative virtual worlds of NVIDIA Omniverse, and beyond.
Learn from some of the world’s brightest minds. Connect with experts. Network with your peers. And discover the technological advancements and groundbreaking research that are making it possible to take on the world’s greatest challenges—together.
March 24, 2022
11:00 am – 11:50 am PDT
Detection of DGA based malicious domain names using real-time ML techniques [S41868]
At the Fall GTC, we talked about advanced threats and how threat actors exploit the changing application landscape. In this session, you'll learn (1) how threat actors use different techniques, specifically algorithmically generated domains (AGDs), to bypass traditional security devices; and (2) about methods of detecting DGAs (domain generation algorithms) using NVIDIA Morpheus-based machine learning model to mitigate threats. Working knowledge of DNS, network security, and basic concepts of data analysis using ML will be helpful. An ensemble of deep learning and traditional ML methods can better detect AGDs in real time and subvert malware attacks that use DGAs. We'll demonstrate end-to-end working prototype of using DNS data, primarily domain names, to detect suspicious domains names potentially generated by DGAs embedded in malicious code running on the compromised systems.
Aditya Sood, Senior Director Threat Research and Strategy, F5
Geng Lin, Executive Vice President and Chief Technology Officer, F5