Is it possible to retrieve ordinary differential equations from observational time series – Application to chaotic systems, and to real world environmental and epidemiological behaviors
Speake: Sylvain Mangiarotti - Cesbio (Toulouse)

Time: 9h30, Thursday, July 11, 2019
Location: Room 611 -612, Building A6, 18 Hoang Quoc Viet.

Abstract: Chaos theory is well designed to tackle the deterministic behaviors that are unpredictable at long term, which is a feature commonly met in real world conditions. Unfortunately, the equations of environmental behaviors are often poorly known. Therefore, it may be useful to develop approaches to obtain or to retrieve such equations directly from observational time series. In the present presentation, after briefly reminding some basis elements of chaos theory, I’ll show that it is possible to retrieve the original Ordinary Differential Equations of numerous theoretical systems, or a good approximation of them when only a few of the original variables are observed. I will then consider several real world problems, focusing in particular on the epidemic of plague that took place in Bombay more than 120 years ago.

Back