Probability Seminar
Short Course : Distributional approximation by Stein’s method
Lecturer: Prof. Robert Gaunt, The University of Manchester, United Kingdom
Abstract: Stein’s method is a powerful (and rather beautiful!) technique for proving quantitative limit theorems in probability with numerous applications throughout the mathematical sciences, including random graph theory, mathematical statistics, and branching processes. In these lectures, we present the basic theory of Stein’s method for normal and Poisson approximation. As an illustration of this theory, we give an elementary proof of the classical Berry-Esseen theorem and derive normal and Poisson approximations for subgraph counts in the Erdos-Renyi random graph model. We conclude by seeing how Stein’s method can be adapted to (many) other limit distributions beyond the normal and Poisson.
Video of the lectures
Day 1 (Thursday 15 October, 2-6 pm): Part 1 - Part 2 - Part 3 - Part 4
Day 2 (Tuesday 20 October, 2-6 pm): Part 1 - Part 2 - Part 3 - Part 4
Cancelled (due to sanitary situation) : Tursday March 19 -- salle "rotule" NO8 (P.2NO8.08) -- Benjamin Arras (Lille 1)