Reinforcement learning of walking behavior for a four-legged robot | IEEE Conference Publication | IEEE Xplore