Title: New Results in Distributed and Combinatorial Optimization Problems in Network Systems
Speaker: Dr. Mayank Baranwal, Post-Doctoral Scholar, Department of EECS, University of Michigan, Ann Arbor.
Time and Date: 10 AM to 11 AM, 29th November, 2019
Venue: IEOR Seminar Room (2nd Floor, IEOR Building)
Abstract: Networks are ubiquitous and often at the heart of fundamental research in science and engineering, from the Internet, to social networks, biochemical reaction networks, transportation networks, power networks, and pharmacology networks that analyze chemical and clinical properties of drug-like molecules. In this talk, I will present some recently developed methods on analyzing such complex network systems, particularly from the point of view of optimization and control theory.
In the first part of the talk, I will describe novel fixed-time convergent algorithms for convex optimization and its application to distributed optimization. Many practical applications, such as economic dispatch in power systems, often undergo frequent and severe changes in operating conditions, and thus require fast solutions irrespective of the initial conditions. Tools from fixed-time Lyapunov theory are leveraged for achieving global fixed-time convergence guarantees in presence of uncertain information and switching topologies.
In the second part of the talk, I will investigate designing entropy regularized algorithms for combinatorial optimization problems. Combinatorial optimization problems arise in many applications in various forms in seemingly unrelated areas such as data compression, pattern recognition, image segmentation, resource allocation, routing, and scheduling, graph aggregation, and graph partition problems. These optimization problems are largely non convex, computationally complex and suffer from multiple local minima that riddle the cost surface. In our work, we are motivated by solutions that are employed by nature to similar combinatorial optimization problems; well described in terms of laws such as minimum free energy principle in statistical physics literature. The resulting approach is independent of initialization, fast and results in high-quality solutions.
Speaker Bio: Mayank Baranwal is a postdoctoral scholar in the Department of Electrical Engineering and Computer Science at the University of Michigan, Ann Arbor. He obtained his Bachelor's in Mechanical Engineering in 2011 from Indian Institute of Technology, Kanpur, and MS in Mechanical Science and Engineering in 2014, MS in Mathematics in 2015 and PhD in Mechanical Science and Engineering in 2018, all from the University of Illinois at Urbana-Champaign. His research interests are in modeling, optimization, control and inference in network systems with applications to distributed optimization, reduction of biochemical networks, transportation networks and control of microgrids.