Learning continuous-action control policies | IEEE Conference Publication | IEEE Xplore