Loading…
This event has ended. Create your own event → Check it out
This event has ended. Create your own
View analytic
Wednesday, April 26 • 10:15am - 10:40am
Building a Platform for Reinforcement Algorithms

Sign up or log in to save this to your schedule and see who's attending!

Q-Learning is a machine learning algorithm which is often a starting point for understanding and developing reinforcement learning concepts. While Q-Learning, itself has significant limits in its application including long learning times and memory constraints for more complex tasks, its theory is widely used for more complex algorithms. Professors who teach reinforcement learning often start with the Q algorithm. One project based method for learning this algorithm is to build a crawler robot that can teach itself to walk. This serves as a great intersection between robotics and machine learning which both have wide interest in hobbyist and academic communities. Market research shows there is a growing number of both robotics hobbyists and people interested in machine learning, however, several barriers exist which make it difficult for hobbyists to know where to begin. Our group is developing a robot kit with fully functional mechanical and electrical systems. Additionally, software that includes example machine learning code and a visualization environment will be developed. This kit will be accessible to hobbyists and professors as a starting point for reinforcement learning. This robot will contribute to reinforcement learning research by making concepts accessible to more people.


Wednesday April 26, 2017 10:15am - 10:40am
417 Mountain View Room, Wilma Sherrill Center

Attendees (1)