论文部分内容阅读
This paper researches robot soccer action selection based on Q learning . The robot learn to activate particular behavior given their current situation and reward signal. We adopt neural network to implementations of Q learning for their generalization pr