2018 March 22 @ LAAS-CNRS, Toulouse, France
This first event will hosts three talks. It is open to any attendees.
9:00 Nicolas Perrin -- How can reinforcement learning exploit the specificities of humanoid robotics?
10:00 Emmanuel Rachelson -- A few topics in Reinforcement Learning (an related ideas).
There are two stories in this presentation and according to the audience's interests we shall invest more time in some parts. I’ll briefly narrate some twists and turns of my short career… and their consequences, in a scientific plot where the guiding thread is the field of Reinforcement Learning and the (current) happy ending is the creation of SuReLI (Supaero Reinforcement Learning Initiative). Then we shall delve into a specific chapter. We shall attempt to understand why, when I had to change my pasty shop, I kept preferring croissants to start off my day, why it is a Mathematical problem with implications in Reinforcement Learning and how it affected my research on learning strategies in stochastic one-player games.
11:00 Justin Carpentier -- Optimal Control and Learning in Robotics - Application to the control of Humanoid Robots
Abstract: In this talk, I will discuss the link between optimal control and machine learning and how they can deal with the control of complex machines like humanoid robots. I will first introduce the fundamental concepts of human and humanoid locomotion. We will see how this locomotion problem can be set up as a complex optimal control problem. In the second part, I will show how we tackle the curse of dimensionality of this optimal control problem by using learning techniques. The last part of this talk will overview some challenges that we are currently facing in Robotics. I will attempt to explicit some directions on the embedding of optimal control into the machine learning formalism. It will be mainly a matter of personal perspectives.