Skip to main content

Deep Learning Malware

Project Description

Using guided learning and reinforcement training techniques for deep analysis of reverse engineered malware to enable similarity analysis and prediction on next malware evolution focused on the adversary tactics modeled for defense actions -  is the goal of Deep Learning Malware

Value Proposition

  • Once the malware is released, other adversaries repurpose and the threat evolves.
  • Pattern matching is static, source code for malware is rare requiring reverse engineering (RE).
  • Advances in RE tools and machine learning have enabled better analysis – including all realms of malware potential.
  • Deep Learning Malware behaviors lead to better/implementable remediations.
  • Enables sharing across energy domains, IT/OT, crowd sourcing and faster understanding of malware targeting energy to develop better indicators and courses of action

Project Objectives

  • Characterized and Eliminate Malware
  • Structured Threat for Visual, Sharable, Actionable, and Implementable (IT/OT)
  • All Possible Paths/Constraints
  • Deep Learning Malware uses recent ML concepts to characterize harder to change, malware behavior in implementable indicators and courses of action

Project Quick Facts

Topic ID: 5.3.2
Status: New

Technical Project Team

  • Lead

    Rita Foster, INL

  • Meng Yue,
  • Jed Haile,
  • Bryan Beckman,
  • Jim Zhan,
  • Jovana Helms,

Project Partners

New York Power Authority
Southern California Edison (SCE)
Detroit Edison

Partner With Us

The Grid Modernization Laboratory Consortium is a strategic partnership between the U.S. Department of Energy and 13 National Laboratories to bring together leading experts and resources. If you would like to partner with GMLC, contact us at the link below.

Contact Us.