An Approximately Optimal Relative Value Learning Algorithm for Averaged MDPs with Continuous States and Actions | IEEE Conference Publication | IEEE Xplore