A Current-Mode Analog Circuit for Reinforcement Learning Problems | IEEE Conference Publication | IEEE Xplore