16 August 2019
Robot Systems Engineering – Introduction to Reinforcement Learning for Robotics
While robot programming can a be complex and time consuming, Machine Learning has the potential to simplify this (and many other) task. In particular, Reinforcement Learning is a powerful approach that enables robots and computers to solve complex decision and control problems using simple reinforcement signals, from playing the game of Go, to teach robots how to grasp objects based on vision.
This course is given by members of the Embodied Systems for Robotics and Learning Unit. You will work together with an team of experienced robotics teachers and researchers to learn how to apply the fundamental techniques of Reinforcement Learning to several robotic problems. The course will cover the most important applications of Reinforcement Learning to robotics, from action planning to movement
Using a wide range of simulation tools and state-of-the-art techniques, you will get hands-on experience on how to solve robotic problems using Reinforcement Learning. The course will cover the underpinning theoretical concepts of reinforcement learning, its potential and limitations. You will have the opportunity to apply these concepts in practical robotic case studies.
EUR 0: Exchange students from a partner university pay no tuition. Guest students pay tuition fees. Tuition vary according to course and your nationality. Read more on official website.