Mini-Course "Probability and Statistics for High Frequency Data " Professor Mark Podolskij

On June 05 - 07 professor of Aarhus University (Denmark) Mark Podolskij read the mini-course "Probability and Statistics for High Frequency Data"

Abstract:

 In this course we will review some recent results on high frequency data of  Ito semimartingales. The notion of high frequency data refers to the sampling scheme where observations are recorded at a very high frequency. Such a setting appears in many applied fields including finance, physics or internet traffic. The aim of the course is to understand the nature of high frequency data from probabilistic and statistical point of view. We will investigate which components of a semimartingale can be identified and provide a complete asymptotic theory for their estimators. From the probabilistic point of view most statistics will have the form of a (generalised) power variation and a large part of the course we be devoted to understanding the limit theory for this type of functionals.

The participants are expected to have standard knowledge about Ito calculus for Brownian motion.