Diffusion gradient temporal difference for cooperative reinforcement learning with linear function approximation | IEEE Conference Publication | IEEE Xplore