ABSTRACT: 
Autonomous robots that can assist humans in situations of daily life
have been a long standing vision of robotics, artificial intelligence,
and cognitive sciences. A first step towards this goal is to create
robots that can learn tasks triggered by environmental context or
higher level instruction. However, learning techniques have yet to
live up to this promise as only few methods manage to scale to
high-dimensional manipulator or humanoid robots. In this tutorial, we
give a general overview on motor skill learning. For doing so, we
discuss task-appropriate representations and algorithms for learning
in robotics. Among the topics are the learning basic movements or
motor primitives by imitation and reinforcement learning, learning
rhytmic and discrete movements, fast regression methods for learning
inverse dynamics and setups for learning task-space policies. Examples
on various robots will be shown; these include ball-paddling,
ball-in-a-cup, robot darts, robot table tennis, learning inverse
dynamics, learning operational space control, and many others.

BIO: 

Jan Peters is a Senior Research Scientist at the Max-Planck Institute
for Biological Cybernetics and head of the new Robot Learning Lab
(RoLL) in the Schoelkopf Department. Before joining MPI, he received a
Ph.D. from the University of Southern California, working at the
Computational Learning and Motor Control lab with Stefan Schaal, Sethu
Vijyakumar and Firdaus Udwadia. He received a M.Sc. in Computer
Science and M.Sc. in Mechanical Engineering from University of
Southern California as well as a Diplom-Informatiker from Hagen
University and a Diplom-Ingenieur in Electrical Engineering from
Munich University of Technology (TU Muenchen). He has been a visiting
researcher at Advanced Telecommunication Research Center (ATR), Kyoto,
Japan in 2000 and 2003, a visiting researcher at National University
of Singapore (NUS) in 2001 and worked as graduate research assistant
at the Institute of Robotics and Mechatronics of the German Aerospace
Research Institute (DLR) in Oberpfaffenhofen, Germany form
1997-2000. His research interests include robotics, nonlinear control,
machine learning, and motor skill learning.

Jan Peters mail(at)jan-peters.net
Max Planck Institute for Biological Cybernetics, D-72076 T\"ubingen, Germany