PHYSTAT seminar: On relating Uncertainties in Machine Learning and HEP

Understanding uncertainties is at the core of data analysis in High Energy Collider Physics. As Machine Learning is rapidly becoming used across the collider physics data analysis pipeline, the High Energy Physics (HEP) community must understand when and what kinds of uncertainties are needed on the predictions of these ML models. At the same time, Uncertainty Quantification is a quickly growing and important topic in ML. In this seminar, we will discuss the connections between uncertainties in HEP and in ML, and progress towards developing and understanding uncertainty quantification for ML in HEP data analysis.
Michael Kagan is a Staff Scientist at SLAC National Accelerator Laboratory.  His research focuses on studying the properties of the Higgs Boson on the ATLAS Experiment at the LHC, and on developing and applying machine learning methods in high energy physics.  Michael received his Ph. D. in physics from Harvard University, and his B.S. in physics and mathematics from the University of Michigan.