Bayesian Machine Learning for Controlling Autonomous Systems

Add to your list(s) Download to your calendar using vCal

• Dr. Marc Deisenroth, Imperial College London
• Wednesday 11 September 2013, 11:00-12:00
• Mech Eng G26.

If you have a question about this talk, please contact Leandro Minku.

Host: Prof. Jeremy Wyatt. Note unusual time and location.

Autonomous learning has been a promising direction in control and robotics for more than a decade since learning models and controllers from data allows us to reduce the amount of engineering knowledge that is otherwise required. Due to their flexibility, autonomous reinforcement learning (RL) approaches typically require many interactions with the system to learn controllers. However, in real systems, such as robots, many interactions can be impractical and time consuming. To address this problem, current learning approaches typically require task-specific knowledge in form of expert demonstrations, pre-shaped policies, or specific knowledge about the underlying dynamics.

In the first part of the talk, we follow a different approach and speed up learning by efficiently extracting information from sparse data. In particular, we learn a probabilistic, non-parametric Gaussian process dynamics model. By explicitly incorporating model uncertainty into long-term planning and controller learning our approach reduces the effects of model errors, a key problem in model-based learning. Compared to state-of-the art RL our model-based policy search method achieves an unprecedented speed of learning. We demonstrate its applicability to autonomous learning in real robot and control tasks.

In the second part of my talk, we will discuss two alternative methods for learning controllers: a) imitation learning and b) Bayesian optimization. In imitation learning, it is no longer necessary to specify a task-dependent cost function. Instead, a teacher demonstrates a task, which the robot should imitate. I will show that probabilistic models are very useful for rapid imitation learning. Bayesian optimization is typically used to optimizes expensive-to-evaluate functions. We successfully applied Bayesian optimization to learning controller parameters for a bipedal robot, where modeling the dynamics is very difficult due to ground contacts. Using Bayesian optimization, we sidestep this modeling issue and directly optimize the controller parameters without the need of modeling the robot’s dynamics.

==== Tea and cookies just after the seminar in the coffee room ====

This talk is part of the Artificial Intelligence and Natural Computation seminars series.

This talk is included in these lists:

• Artificial Intelligence and Natural Computation seminars
• Computer Science Departmental Series
• Computer Science Distinguished Seminars
• Mech Eng G26
• computer sience

Note that ex-directory lists are not shown.