ML & physics, credit for the graphics Kyle Cranmer

Our scientific interest is profoundly grounded in physics and concerns fundamental questions about the use of algorithms to solve both applied and academic problems.

The ongoing work and interest concerns understanding why and how deep learning and other machine learning methods work, how to improve them further and adapt them to emerging applications for instance in physics.

Our research span a range of problems that can be defined as the Statistical Physics of Complex Systems. Methodologically we adapt and develop tools of statistical physics of glassy and disordered systems to describe high-dimensional spaces, probability distributions and landscapes that originate from interactions of many correlated elements with applications spannin