Jan Hendrik Metzen Frank Kirchner Mark Edgington Yohannes Kassahun Towards efficient online reinforcement learning using neuroevolution. 1425-1426 2008 GECCO http://doi.acm.org/10.1145/1389095.1389371 conf/gecco/2008 db/conf/gecco/gecco2008.html#MetzenKEK08