Internship and thesis proposals
Deep learning of scale free processes

Domaines
Statistical physics
Nonequilibrium statistical physics
Non-equilibrium Statistical Physics

Type of internship
Théorique, numérique
Description
In the statistical physics context many problems of interest, related to critical phenomena are associated to fractal structures. This problem is notoriously difficult with standard ML. Taking the percolation problem as a study case the task will consist to regress from images of percolation states the size of the maximal cluster and whether it is a spanning cluster via new ML algorithms. A second and related task would be to consider different 2D conformal field theory, characterized by some fractal exponents ot their central charge.
Contact
Sergio Chibbaro
Laboratory : LISN -
Team : Decipher/TAU
Team Website
/ Thesis :    Funding :