Developing agent models with a neural reinforcement technique | IEEE Conference Publication | IEEE Xplore