Enrolment options

This course is an introduction to mathematical statistics and statistical learning. Students learn to choose a statistical model, estimate its parameters, build confidence regions and tests, and compare these procedures with one another: decision theory, the Cramér-Rao bound, optimality of tests and asymptotic theory. The last three lectures are devoted to supervised learning: the Bayes rule, model-based approaches, empirical risk minimisation, PAC theory and the Vapnik-Chervonenkis dimension. The notions are put to work in exercise classes on concrete examples (signal and image processing, econometrics, environmental sciences, pattern recognition) and in Python through homework assignments and numerical explorations.

Guests cannot access this course. Please log in.