"Learning to overtake in TORCS using simple reinforcement learning."

Daniele Loiacono et al. (2010)
a service of Schloss Dagstuhl - Leibniz Center for Informatics