Official description

This course will cover conceptual foundations and practical applications of a number of common statistical learning methods – from classical tools to their modern incarnations, with the emphasis on the connection among the different tools and their relation to a number of models of mathematical physics.

Specific topics will include: maximum likelihood estimation, Bayesean methods, clustering and dimensionality reduction, neural networks and Boltzmann machines, supervised and unsupervised learning, feature extraction and learning.

The mathematical methods will be illustrated by examples from non-equilibrium statistical mechanics, physical chemistry, population dynamics and epidemiology, quantum optics, solid state physics, chemical reactions and gene regulation.