Statistical Inference Under Information Constraints: User level approaches

Abstract:

In this talk, we will present highlights from some of the work we have been doing in distributed inference under information constraints, such as privacy and communication. We consider basic tasks such as learning and testing of discrete as well as high dimensional distributions, when the samples are distributed across users who can then only send an information-constrained message about their sample. Of key interest to us has been the role of the various types of communication protocols (e.g., non-interactive protocols vs interactive protocols, etc). We will also discuss some recent results on estimation of discrete distributions with user level information constraints, where multiple samples per user are available at each user.

This talk is based on works with Clement Canonne, Yuhan Liu, Ziteng Sun, and Himanshu Tyagi.

Bio:

Jayadev Acharya is an associate professor in Electrical and Computer Engineering at Cornell University. He has a PhD from UCSD and was a postdoc at MIT. He is broadly interested in problems at various intersections of information theory, statistical inference, algorithms, and machine learning. He received a best paper award at ALT 2020, a best paper honorable mention at ICML 2017, and the Jack Wolf student paper award at ISIT 2010.