Scalable reward learning from demonstration | IEEE Conference Publication | IEEE Xplore