In this paper we introduce four new learning models: impulse balance learning, impulse matching learning, action-sampling learning, and payof-sampling learning. With this models and together with the models of self-tuning EWA learning and reinforcement learning, we conduct simulations over 12...