Title: On data-driven forecasting of financial time series based on SDEs with memory

Abstract: In this talk we discuss some approaches for non-linear data driven modeling and forecasting of long memory time series that particularly appear in the context of financial applications. Our approach is based on SDEs that, with general coefficient functions, provides a rich and flexible model class. On top of theoretical results, we illustrate the applicability of our approach with simulation study. Our approach is based on an interplay between analysis, probability, statistics, and machine learning. 

This talk is based on a joint work with Mahdi Dehshiri and Kerlyns Martinez.

--

Per coloro che volessero collegarsi da remoto ecco di seguito il link: