Collaborative Decision-Making Under Adversarial and Information Constraints
Centre for Networked Intelligence, IISc via YouTube
Overview
Learn about collaborative decision-making challenges in a lecture from Prof. Aritra Mitra of North Carolina State University that explores how to optimize learning from multiple agents while dealing with adversarial threats and communication constraints. Examine the linear stochastic bandit framework to understand the balance between beneficial collaboration and potential disruption from adversaries, with insights into robust algorithms that effectively manage the exploration-exploitation tradeoff. Discover how rate-limited channels affect linear stochastic bandits and learn about novel adaptive encoding strategies that achieve optimal regret with minimal communication requirements. Gain valuable insights from Prof. Mitra's expertise in control theory, optimization, and machine learning, developed through his distinguished academic career at institutions including the University of Pennsylvania, Purdue University, IIT Kanpur, and Jadavpur University.
Syllabus
Time: 5:00 to PM
Taught by
Centre for Networked Intelligence, IISc