Prof. Panos Patrinos, PhD
KU Leuven, Dept. Electrical Engineering (ESAT), STADIUS Center for Dynamical Systems, Signal Processing and Data Analytics
Title of the invited lecture: Algorithms for large-scale structured nonconvex optimization
Optimization problems are ubiquitous in many engineering and science disciplines, such as machine learning, signal processing, data science, communications, control and robotics. Optimization methods have to cope well with the demanding requirements of modern applications. These include handling large numbers of variables and constraints, being amenable to distributed computations and finally being able to cope with nonconvexity, typically encountered in applications such as deep learning. In fact, it is has long been recognized that the great watershed in optimization is between convexity and non convexity. However, in recent years important progress has been made in bridging these two fields. On one hand, it has been recently discovered that certain structured nonconvex problems posses benign landscapes which means they are easier to solve to global optimality. On the other hand, numerical algorithms that have been traditionally developed under the realm of convexity, have had recently many successes in nonconvex optimization. In this talk, I will present highlights of these achievements and focus on our recent work on provably convergent algorithms for nonconvex, nonsmooth, large-scale optimization, as well as ways of accelerating them. The key enabler for these advancements is the concept of proximal envelopes. Examples coming from the fields of machine learning, signal processing and control of autonomous systems will be presented.
Short CV: Panagiotis (Panos) Patrinos is currently associate professor at the Department of Electrical Engineering (ESAT) of KU Leuven, Belgium. In 2014 he was a visiting professor at Stanford University. He received his PhD in Control and Optimization, M.S. in Applied Mathematics and M.Eng. from the National Technical University of Athens in 2010, 2005 and 2003, respectively. After his PhD he held postdoc positions at the University of Trento and IMT Lucca, Italy, where he became an assistant professor in 2012. His current research interests lie in the intersection of optimization, control and learning. In particular he is interested in the theory and algorithms for large-scale structured nonconvex optimization as well as learning-based control with a wide range of applications including autonomous vehicles, machine learning, signal processing and energy. He is the author of more than 90 peer-reviewed publications and recipient of the IJCTA best paper award 2020. He has served as general chair of the 4th European conference on Computational Optimization and the 38th Benelux meeting on Systems and Control.