User:Quaenuncabibis/Clément Hongler

Clément Hongler (born 1985 in Geneva) is a Swiss mathematician and physicist. He is a professor of mathematics at EPFL (École Polytechnique Fédérale de Lausanne) and the head of the Chair of Statistical Field Theory.

Career
Clément Hongler received his M.Sc. in mathematics from the EPFL in 2008. Under the direction of Stanislav Smirnov, he earned his Ph.D. in 2010 at the University of Geneva for his thesis on "Conformal Invariance of Ising Model Correlations." From September 2010 to July 2014, he worked as a Ritt Assistant Professor at Columbia University.

He has served as the Chair of Statistical Field Theory (CSFT) at EPFL since July 2014, first a an assistant professor and since 2018 as associate professor.

Research
Statistical mechanics, quantum field theory, learning theory, and decentralized systems make up the majority of our study.

His current work focuses on lattice models and their relationships to learning theory, statistical mechanics, quantum field theory, decentralized systems, and quantum field theories.

the examination of the relationship between conformal field theories and lattice models. More specifically, we are interested in exposing conformal field theory structures within lattice models, rigorously explaining phase transitions in lattice models in terms of conformal field theories, and connecting field theories with probabilistic objects like random curves and fields. Our goal is to properly combine discrete and continuous models, statistical and quantum theories, algebraic and probabilistic structures that appear in the study of phase transitions by building mathematically precise bridges between these items. the dynamics of learning, especially during training of deep neural networks. More specifically, we use methods from probability, functional analysis, and algebra to study the dynamics of neural networks during supervised learning (such as regression or classification) or unsupervised learning (such as generative adversarial networks). We specifically look into how neural networks relate to other learning strategies (such kernel methods) and what makes them so effective for a variety of tasks.

Blockchain technology in particular, along with algorithmic game theory, are examples of decentralized systems.

Distinctions
Hongler received the 2014 Blavantik Awards for Young Scientists and the 2017 Latsis University Prize.