October 3, 2024 – Lecture: From a New Approach to Viral Propagation to Predictions in Climatology
Alain OUSTALOUP
October 3, 2024 Lecture Hall 36.02 – 11:00 a.m.
Professor Emeritus, ENSEIRB-MATMECA – Bordeaux INP / IMS Laboratory
Biography of Alain Oustaloup – A 1973 graduate of ENSEIRB, Alain Oustaloup is currently professor emeritus at ENSEIRB-MATMECA – Bordeaux INP. After synthesizing complex non-integer differentiation and integration, and then resolving the stability-accuracy dilemma in control theory and the mass-damping dilemma in mechanics, he invented the CRONE control system and the CRONE suspension. More recently, he has done it again by proposing a new non-integer-power predictor, which he applies in epidemiology and climatology. His work was recognized with a CNRS Silver Medal in 1997 and a Grand Prize from the Academy of Sciences in 2011.
Conference Abstract: Recognized as a “CNRS Highlight of 2021,” the FPM (Fractional Power Model)—which generalizes linear regression—is a three-parameter model in which the power alone indicates the progression, stabilization, or regression of an epidemic. Through its non-integer power, this model contributes to unifying diffusion phenomena in physics and viral propagation in epidemiology. Its ability to accurately represent real-world data stems from its capacity to account for an unlimited number of internal dynamics with varying rates. The model thus makes it possible to represent all the internal dynamics of an epidemic, from the slowest—found in highly depopulated rural areas—to the fastest—found in densely populated major cities. Its predictive power stems from its ability to account for the entire past by weighting it appropriately. The model is, in fact, equipped with a long-memory predictive form, which means that every predicted value is a function of all past values—values that are favorably weighted according to a forgetting factor (which is not unlike a subtle form of memory). The model thus offers the advantage of making the best use of the past, especially since only the past can be used to predict the future—a predictive characteristic that makes this model a reliable tool for decision-makers. The model’s representativeness has been validated using official data from the French Ministry of Health on the spread of COVID-19, particularly the time series of infections and hospitalizations. Its predictive power has also been validated by verified predictions during lockdowns and vaccination campaigns, and even for the vaccination process itself.
Finally, the model is further validated by verified predictions in its favor, conducted in the field of climatology through the evolution of atmospheric CO2 concentrations and rising average sea levels.