A story of least favorable distributions
ECE 595 Seminar Series
November 25, 2024
11:00 AM - 12:00 PM
A story of least favorable distributions
Speaker: Alex Dytso, Qualcomm
Abstract: This talk provides an overview of distributions that naturally arise in estimation theory and information theory, which can loosely be termed "least favorable." Drawing from Dytso's journey and experiences in these fields, the lecture will explore basic but important problems, such as determining the capacity of amplitude-constrained channels in information theory and identifying distributions that induce linear estimators in estimation theory. A key focus will be on demonstrating the deep connections between these seemingly disparate problems. The session will conclude with open problems and those that Dytso is working on in industry.
Speaker bio: Alex Dytso received his PhD degree from the University of Illinois Chicago in 2016. From September, 2016, to August, 2020, he was a post-doctoral associate with Princeton University's department of electrical engineering From 2020 to 2022, he was an assistant professor of electrical and computer engineering at the New Jersey Institute of Technology. Currently, Dytso is a staff engineer with Qualcomm Flarion Technologies Inc. His current research interests include the areas of multi-user information theory and estimation theory and their applications in wireless networks.
Faculty host: Daniela Tuninetti, danielat@uic.edu
Date posted
Nov 25, 2024
Date updated
Nov 27, 2024