Abstract: 

Reinforcement learning is getting a lot of attention lately. People are excited about its potential to solve complex problems in areas such as robotics and automated driving, where traditional control methods can be challenging to use. In addition to deep neural nets to represent the policy, reinforcement learning lends itself to control problems because its training incorporates repeated exploration of the environment. As such exploration is time-consuming and costly or dangerous when done with actual hardware, a simulation model is often used to represent the environment. In this talk, we provide an overview of reinforcement learning and its application to teaching a robot to walk. We discuss the differences between reinforcement learning and traditional control methods.
 
Specific topics of reinforcement learning covered in this presentation include:
- Creating environment models
- Crafting effective reward functions
- Various approaches to training policies
- Deploying to embedded devices through automatic code generation for CPUs and GPUs

Bio: 

Craig Buhr received his B.S., M.S. and Ph.D. from the School of Mechanical Engineering at Purdue University. He joined MathWorks in 2003 as a Senior Software Engineer for the Controls and Identification products. He currently manages the engineering team responsible for the control design products. (Control System Toolbox, Robust Control Toolbox, Model Predictive Control Toolbox, Simulink Control Design and Reinforcement Learning Toolbox).
The team is focused on designing and developing software tools that enable engineers to efficiently design and analyze control systems for industrial applications.  His research interests include dynamic system modelling, control theory, machine learning, reinforcement learning and computer aided control system design.