Skip to Main Content
College Home Page
E C E Home Page

EE Seminars

Data Science of Intelligent Interconnected Systems


  Add to Google Calendar
Date:  Fri, March 11, 2022
Time:  9:30am - 10:30am
Location:  Holmes 389; online available, see below registration info
Speaker:  Igor Molybog, Ph.D. candidate, University of California, Berkeley

Online available, register for connection info at https://forms.gle/yeGtuLSFYqgbEJg86

Abstract

The solutions of the major problems of the XXI century require novel computational techniques for analytics of the data of complex safety-critical systems. A prominent large-scale example of such a system that I was working on throughout my Ph.D, is the sensor network in the control loop of an electrical power grid. It is known that the associated generic problem of recovery of the state of a power system is computationally difficult yet critical to be solved efficiently. The three main paths to reducing the complexity of computation associated with a sensor network are to increase the density of sensors, to place them strategically, or to enhance their security and accuracy to bound the properties of possible errors and faults of measurement. We report our investigation of all three of these paths. We provide numerical quantification to the computation complexity of inverse problems depending on the total number of measurements, the properties of the graph structure of the measurement network, and the signal to noise ratio expressed as the ratio of good and bad measurements. We address both the computational complexity and the matter of constructing a suitable efficient algorithm in each of the three scenarios with guaranteed performance. Being highly general, our theoretical results offer implications that should be considered on multiple stages throughout the life cycle of a non-linear measurement system, starting from the design of the sensing mechanisms, to ensure robustness and security of operation both in normal conditions and in situations of an emergency such as during a cyber-attack.

 

Bio

I received a bachelor's of science degree in Applied Physics and Mathematics from the Moscow Institute of Physics and Technology in 2017, where I have worked in the Intelligent Data Analytics lab on NLP and traffic analytics projects. Now, I am a Ph.D. candidate at the department of Industrial Engineering and Operations Research at University of California, Berkeley working with Professor Javad Lavaei on problems of algorithmic analysis and optimal control of complex safety-critical systems, such as power systems, transportation and telecommunication networks, AI recommendation and navigation systems, robotic systems and others. My research spans the theory of non-convex and conic optimization, stochastic control, machine learning, and computational and sampling complexity of learning algorithms. I aim to design data processing algorithms which are robust to noise and highly scalable with the amount of available computational resources.

Return to EE Seminars