One Size Does Not Fit All: A Game-Theoretic Approach for Dynamically and Effectively Screening for Threats

Speaker: 
Arunesh Sinha
Date/Time: 
Wednesday, May 20, 2015 - 14:00
Location: 
RTH 306
An effective way of preventing attacks in secure areas is to screen for threats (people, objects) before entry, e.g., screening of airport passengers. However, screening every entity at the same level may be ineffective and undesirable. The challenge then is to find a dynamic approach for screening, allowing for more effective use of limited screening resources, leading to improved security. We address this challenge with the following contributions: (1) a threat screening game (TSG) model for screening domains; (2) an NP-hardness proof for computing the equilibrium of TSGs; (3) a scheme for decomposing TSGs into subgames to improve scalability; (4) a column generation approach to solve TSGs which includes a novel multidimensional knapsack slave formulation and heuristics for faster computation; and (5) a minimax regret-based trade-off analysis for handling uncertainty in the number of screenees and choosing a robust screening strategy. This talk will focus more on the model and we welcome any feedback that could improve the model or algorithmic techniques used in this ongoing work.

Bio: Dr. Arunesh Sinha is a postdoctoral scholar with Prof. Milind Tambe at the Computer Science Department of University of Southern California. He received his Ph.D. from Carnegie Mellon University in Aug 2014, where he was fortunate to be advised by Prof. Anupam Datta. He obtained his undergraduate degree in Electrical Engineering from IIT Kharagpur in India. He has industry research experience in form of internships at Microsoft Research, Redmond and Intel Labs, Hillsboro. He was awarded the Bertucci fellowship at CMU in appreciation of his novel research.

Dr. Sinha has conducted research at the intersection of cyber-security, machine learning and game theory. He introduced a novel game theoretic model of auditing for enforcement of policies in large organizations. He has also worked on the use of machine learning to learn and enforce access policies. His interests lie in the theoretical aspects of multi-agent interaction, machine learning, security and privacy, along with an emphasis on real-world applicability of the theoretical models