Séminaire MODAL'X : Camille Frévent (Université de Lille)

Abstract :

In recent years, spatial analysis of functional data has received growing interest.
This talk will be divided into two parts, each focusing on a different aspect of spatial analysis of functional data.

The first part will introduce a novel approach to spatial cluster detection of functional data.
This approach is based on the use of spatial scan statistics, and is particularly useful in environmental surveillance: Given the known adverse health effects of pollutants, it is crucial to identify and locate environmental black spots.
Three new test statistics have been proposed and evaluated through a simulation study. These statistics were then applied to detect environmental black spots in northern France.

Moving onto the second part of the talk, a new approach to the autoregressive spatial functional model will be presented.
This approach is based on the notion of signature, which represents a function as an infinite series of its iterated integrals and can be applied to a wide range of processes.
The estimation problem of the model will be discussed, and the new approach will be compared to the traditional autoregressive spatial functional model through a simulation study.