Q-Learning with Recurrent Neural Networks as a Controller for the Inverted Pendulum Problem.

Ahmet Onat, Hajime Kita, Yoshikazu Nishikawa

Browse the full ICONIP paper archive.