Exploiting Domain Symmetries in Reinforcement Learning with Continuous State and Action Spaces | IEEE Conference Publication | IEEE Xplore