The last stream of our research is devoted to the development of novel game theoretic and robust models and solution approaches to address address real-world large scale resource allocation problems in the face of strategic adversaries. It is motivated by resource allocation problems arising in public safety and security, and in biodiversity preservation.
Part of this work is inspired from discussions with officials at the Transportation Security Administration and by the effectiveness/efficiency trade-offs that they face given limited resources. For example, we have proposed a tractable framework for adaptively screening for threats at airport checkpoints when the passenger arrival times are uncertain. We have also studied the problem of strategically allocating Federal Air Marshals on flights in the presence of scheduling constraints.
The remainder of our work is motivated by issues related to security surrounding technology. For example, we have proposed game theoretic models that network administrators can employ to deceive cyber adversaries by strategically camouflaging certain components of the computer network. Also, motivated by the emerging application of utilizing mobile sensors (e.g., UAVs) for patrolling, we have proposed a game theoretic model which enables the joint allocation of human patrollers and mobile sensors to defend potential targets (e.g., national parks). This work is currently supported by a five year $6 million MURI (Multidisciplinary University Research Initiative) U.S. Army Research Office grant.
Cyber camouflage games for strategic deception
(*) O. Thakoor, M. Tambe, P. Vayanos, H. Xu, C. Kiekintveld, F. Fang
In Proceedings of the 10th International Conference, GameSec, 2019.
The price of usability: designing operationalizable strategies for security games
(**) S. M. Mc Carthy, (*) C. M. Laan, K. Wang, P. Vayanos, A. Sinha, and M. Tambe
In Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI), 2018.
Deceiving cyber adversaries: a game theoretic approach
A. Schlenker, M. Tambe, L. Tran-Thanh, P. Vayanos, Y. Vorobeychik, O. Thakoor, H. Xu, and F. Fang
In Proceedings of the 17th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2018.
Equilibrium refinement in security games with arbitrary scheduling constraints
K. Wang, Q. Guo, P. Vayanos, M. Tambe, and B. An
In Proceedings of the 17th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2018.
Strategic coordination of human patrollers and mobile sensors with signaling for security games
H. Xu, K. Wang, P. Vayanos, M. Tambe
In Proceedings of the 32nd AAAI Conference on Artificial Intelligence, 2018.
(**) S.M. Mc Carthy, P. Vayanos and M. Tambe
In Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence (IJCAI), pp. 3770-3776, 2017.
U.S. Army Army Research Office, MURI (Multidisciplinary University Research Initiative)
Role: PI (since August 2019; original PI and project lead: M. Tambe)
Award ID: W911NF-17-1-0370
Total Award Period Covered: 05/16/2017-05/15/2022
Total Award Amount: $6,206,947
Own Share: $313,353
DARPA Seedling (via Lockheed Martin Corporation)
Role: Co-PI (PI: M. Tambe, Co-PI: B. Dilkina)
Funded for the period: 12/18/2017-1/07/2018
Total award amount: $100,000
U.S. Army Research Laboratory, Army Research Office
Role: Co-PI (joint with M. Tambe and E. Rice)
Award #: W911NF-11-1-0332
Funded for the period: 2/1/2017-9/1/2017
Total award amount: $500,000
Own share: $150,000