Alexander Barg

I am a Professor in the Department of ECE and Institute of Systems Research of the University of Maryland, with affiliate appointments in the Departments of Computer Science and Mathematics. I am also a faculty member in the Program of Applied Mathematics and Scientific Computing of UMD.

Research: I am broadly interested in mathematical problems of information and coding theory, related problems in distance geometry and combinatorics, as well as random models of managing and processing information. I have worked on asymptotic estimates of code parameters in various metric spaces, coding problems for digital fingerprinting, random codes and error exponents, estimates of the size of spherical codes with few distances, combinatorics of association schemes of coding theory, and more recently on algebraic code constructions for various models of distributed storage, private distribution estimation, polar codes, and a variety of other problems.

Currently I am interested in problems related to discrepancy in the Hamming space and other spaces and its connections to various topics in computer science (uniform distributions, derandomization, approximations and VC dimension). Another recent area of research concerns data representation on graphs (finite and infinite) wherein the contents of a vertex is determined by the values of its neighbors in the graph. I also maintain interest in some of the earlier problems such as algebraic codes for distributed storage (the locality and regeneration problems).

More details about my research appear on the Research and Publications pages accessible through the menu on this page and on my GoogleScholar profile.

Teaching: A few years ago I taught a special topics course on "Modern discrete probability" and more recently another topics course on "Classical and Quantum Codes" (co-taught with Victor Albert), both cross-listed between ECE, CS, and Math. I regularly teach graduate courses in Random Processes, Information Theory, and well as undergraduate classes.