Probability and Stochastic Analysis Seminar  RSS

25/01/2023, 16:00 — 17:00 — Online
Milton Jara, Instituto de Matemática Pura e Aplicada

Stein's method for Markovian martingales

Stein's method is an increasingly popular way to derive quantitative versions of weak convergence theorems, like the central limit theorem. In this talk we use Stein's method to derive a quantitative CLT for Dynkin martingales of Markov chains. Despite its simplicity, we show with some examples that the bounds we obtain in the context of interacting particle systems are surprisingly sharp.

For more info: https://spmes.impa.br

link

Except for a few of the oldest sessions these are from the Seminário de Probabilidade e Mecânica Estatística at IMPA which is co-sponsored by several institutions, in particular Instituto Superior Técnico.