Our conference paper on inverse reinforcement learning under information asymmetry constraints in collaboration with Lockheed Martin is on arXiv! We investigate how privileged information allows simultaneous identification and mitigation of adversarial entities. We draw on tools from revealed preference in economics, and principal agent problem in contract theory. We illustrate our algorithm’s performance in a cognitive radar scenario.
This manuscript is under review at ICASSP, 2023. Journal version coming soon!