Post by The University of Texas at San Antonio

What if AI could learn more from failure than success? Professor Yongcan Cao from the Margie and Bill Klesse College of Engineering and Integrated Design at UT San Antonio is challenging traditional AI training with On-F (On-Policy Reinforcement Learning from Failure), an approach that teaches autonomous systems to learn by avoiding past mistakes instead of relying solely on expert demonstrations. Supported by a $500,000 Office of Naval Research grant, this work is making AI training more efficient while advancing applications in robotics, autonomous vehicles, healthcare and disaster response. Learn more about Dr. Cao's groundbreaking research and its impact on the future of autonomous systems: https://bit.ly/44FQ3yk