Colloquium – Konstantin Matchev – Interplay of Fundamental Physics and Machine Learning

Speaker: Konstantin Matchev (University of Florida)

Title: Interplay of Fundamental Physics and Machine Learning

Abstract:

The recent boom in AI algorithms and their successful application to high-dimensional large datasets from all aspects of life motivates us to approach a number of problems in physics and astronomy as machine-learning tasks, with custom loss functions and specialized machine-learning architectures. In this talk I will demonstrate several applications of this new machine learning approach: for the discovery and identification of continuous symmetries in the data, for building true and beautiful theoretical models, for deriving the optimal variables for experimental data analyses, and for obtaining novel analytical formulas with symbolic regression.

Host: Sergei Gleyzer