Reinforcement Learning Robots in Partially Observable Environments

Schmidhuber, Juergen
Reinforcement Learning Robots in Partially Observable Environments
Journal papers and conference papers (HTML - 100 KB)

Abstract: Robot learning in realistic environments requires novel algorithms for learning to identify important events in the stream of sensory inputs, and to temporarily memorize them in adaptive, dynamic, internal states until the memories can help to compute proper control actions. We present a list of selected recent publications on robot learning / soccer learning.