Robot Learning in Practice [From the Guest Editors

摘要

Robotics researchers have made great efforts to model real environments in designing controllers. Thanks to the rapid increases in the computational speed of affordable computers and to the developments in sophisticated machine-learning algorithms, the acquisition of an environmental model and designing controllers from measured data with less prior information has become a practical approach. Work toward developing such robot-learning approaches can be expected to significantly contribute to the entire field of robotics research.