Jean Barbier
Associate Research Officer - Abdus Salam International Center for Theoretical Physics (ICTP)
Jean Barbier is an Associate Research Officer at the Abdus Salam International Center for Theoretical Physics (ICTP) in Trieste. He received his PhD in 2015 from the Ecole Normale Superieure (ENS) of Paris where he applied statistical physics as well as message-passing algorithms to probabilistic graphical models, with applications ranging from signal processing to combinatorial optimization or coding theory.
He then joined the EPFL in Lausanne as a Research Scientist from 2015 until 2018, with a focus on the rigorous aspects of high-dimensional statistical problems. He then occupied the Junior Professor Chair in Data Science at the ENS of Paris before joining the ICTP.
His main research interests are in statistical physics and its interdisciplinary connections with statistical inference, information theory and machine learning.
How would you define your field of study? What is your vision about it and which are the topics you're most passionate about?
Statistical physics of disordered systems, and its interdisciplinary connections with information theory, high-dimensional statistical inference and machine learning.
How do you expect your experience in IMM to be? Why did you accept to teach for this project?
I have no expectations other than an extremely enriching experience, both culturally and scientifically.
Do you have one of two favorite quotes you would like to share and/or a personal “motto”?
An expert is a man who has made all the mistakes which can be made, in a narrow field. - Niels Bohr
Lectures
Beyond the thermodynamic limit in high-dimensional inference, 40 years of Replica Symmetry Breaking, Sapienza University (Rome), 2019
0-1 Phase Transitions in High Dimensional Inference , SAMSI Institute, 2019
Mutual information for the stochastic block model - a direct proof, London Symposium on Information Theory, 2019
More Lectures