Countering the Ability to Trick Machine Learning

Army Cyber Institute Photo Gallery

Keyword:

Category:

Tag:

Sort By:

Countering the Ability to Trick Machine Learning

With society becoming dependent on Machine Learning for things like facial recognition to optimizing power generation, there is significant research being conducted on how to cause the learning algorithms to produce undesirable results using falsified or tainted data. While this research is becoming more understood, the research on countering these types of attacks is not as advanced. During this projects, ACI researchers taught cadets basic Machine Learning principles, exposed to a number of different techniques used to fool Machine Learning and asked to participate in investigative research on countering these techniques.

Related Imagery

Cadet Cyber 9/12 Policy Team wins 2d place at Columbia University event

Joint Service Academy Cybersecurity Summit 2023

Col. Jeff Erickson, The Army Cyber Institute at West Point Chief of Staff, gives a keynote presentation on building #cyber resiliency of public and private infrastructure at Cyber Defense Workshop Halfway Across the World - (CYDEF) in Tokyo, Japan. The workshop brought together multi-national partners to address the theme of ‘Assessing National Risk and Establishing Multi-Stakeholder Cooperation in Cyber Defense.

Hackathon - West Point Cadet Hunter Hill and Capt. Mark Lesak, Center for Mechanical Engineering instructor and Army Cyber Institute researcher, work together to design and provide technical solutions to challenges for equipment used by USASOC. USASOC technology developers, West Point faculty from Center for Innovation and Engineering and the Robotics Research Center, staff from the Army Cyber Institute, and Cadets teamed up to pursue a rapid design and prototyping effort.

Pulling an all-nighter - Members of the Army Cyber Institute demonstrate the capabilities of their High Frequency technology and equipment to Mr. Adam Nucci, HQDA Cyber, during Cyber Blitz.

Capt. Victor Mitsuoka and Lt. Col. Brett Lindberg use a combination of RTK GPS receivers and geodetic benchmarks to more accurately provide locations of targets within centimeters on augmented reality overlays.

Melding of the Minds - The ACI CSMO team hosts an informal High Frequency Communications seminar with four subject matter experts: Al Danis, John Rosica, Jon Bentley, and Rob "Bob" Melville. They conducted HF radio demonstrations and discussed antenna tuning research.

Army Cyber Institute
180913-A-ZZ999-1001.PNG

DOWNLOAD PHOTO
(0.18 MB)

Tags
USMA, research, ACI, Army, Machine Learning, tensorflow

Back to Main Gallery