-
Center of Autonomy and AI, Center for Cyber-Physical Systems and the Internet of Things, and Ming Hsieh Institute Seminar Series
Wed, Jan 26, 2022 @ 02:00 PM - 03:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Dhruva Kartik Mokhasunavisu, Electrical Engineering, University of Southern California
Talk Title: Stochastic Zero-sum Games between Two Competing Teams
Series: Center for Cyber-Physical Systems and Internet of Things
Abstract: Intelligent autonomous systems can be employed in many areas such as transportation, power grids, manufacturing, wildlife conservation etc. Agents in these systems act sequentially over time in uncertain and dynamically evolving environments. In many cases, information about the system is dispersed among the agents and hence, there is asymmetry in the information used by the agents to select their actions. Furthermore, the agents involved may have different (possibly conflicting) objectives. Due to this asymmetry in information and objectives, agents must act in a decentralized and strategic manner. The focus of this talk is on adversarial interactions that can be modeled as stochastic zero-sum games between two teams of agents. In this setting, agents within a team are cooperative. However, the team as a whole is non-cooperative and adversarial with respect to the other team. Finding the value and Nash equilibria of games with incomplete and asymmetric information is notoriously hard. For our game model, we provide a dynamic programming characterization of the value (if it exists). If the value does not exist, the dynamic program provides us with bounds on the upper and lower values. For some specialized game models, we also characterize a min-max strategy for the minimizing team. Further, we propose a computational methodology to solve the dynamic program and illustrate it with the help of an example.
Biography: Dhruva Kartik is a Postdoctoral Researcher at University of Southern California (USC), Los Angeles co-advised by Prof. Rahul Jain and Prof. Pierluigi Nuzzo. He received his Ph.D. in electrical engineering from USC in 2021, where he worked with Prof. Urbashi Mitra and Prof. Ashutosh Nayyar. While he was at USC, he was an Annenberg Fellow from 2015 to 2019. He received his Bachelor of Technology degree in electronics and communication engineering from the Indian Institute of Technology Guwahati (IITG), India, in 2015. His research interests include multi-agent systems, game theory, probabilistic contracts, and reinforcement learning.
Host: Pierluigi Nuzzo, nuzzo@usc.edu
Webcast: https://usc.zoom.us/webinar/register/WN_zyIBh_1gQLmKpMJG0GyLxwLocation: Online
WebCast Link: https://usc.zoom.us/webinar/register/WN_zyIBh_1gQLmKpMJG0GyLxw
Audiences: Everyone Is Invited
Contact: Talyia White