Loading…
Back To Schedule
Monday, November 9 • 12:40pm - 1:10pm
Train but Verify: Towards Practical AI Robustness

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Dr. Van Houdnos will describe his research into developing a continuous integration/continuous delivery Fast AI Test Harness (FAITH) to verify the robustness properties of the system’s machine learning components. The objective of the Train, but Verify project is to: (1) research and refine an AI Engineering process designed to train AI systems to have specific robustness properties, and (2) develop a testing framework capable of verifying these robustness practices. This presentation will be followed by a live Q&A period with John Helland from the project research team.

Mr. Helland and Ms. Smith will moderate a live Q&A session following Dr. Van Houdnos's presentation.

Speakers
avatar for Ms. Tanisha Smith

Ms. Tanisha Smith

Technical Manager, CMU SEI
avatar for Mr. Jon Helland

Mr. Jon Helland

Associate Machine Learning Researcher, CMU SEI
avatar for Dr. Nathan Van Houdnos

Dr. Nathan Van Houdnos

Senior Machine Learning Research Scientist, CMU SEI
Learn more about Nathan at https://resources.sei.cmu.edu/library/author.cfm?authorid=484947. Contact him via https://www.sei.cmu.edu/contact-us/index.cfm?f=Nathan&l=VanHoudnos


Monday November 9, 2020 12:40pm - 1:10pm EST
Virtual