PAC LEARNING FOR STRATEGIC CLASSIFICATION

REGISTER HERE
Zoom link and SEC watch location will be shared with those who register in advance.

 

The celebrated field of mechanism design studies how a system designer can design agents' incentives, and consequently their actions, in order to steer their joint decisions towards a desirable outcome. This talk also examines the intervention of agents' actions but through a fundamentally different yet equally important "knob" --- i.e., influencing agents' decisions by designing the available information to each agent. This task, a.k.a. information design, is particularly relevant in this digital era with unprecedented convenience in  accessing information today and has found numerous applications. This talk will provide a thorough algorithmic examination for a foundational model in this space, namely the Bayesian Persuasion model. We will discuss its computability, learnability and broad applicability to domains of societal importance including security and public safety, sustainability and traffic control.