Abstract: We introduce a diffusion-based algorithm in which multiple agents cooperate to predict a common and global state-value function by sharing local estimates and local gradient information ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results