FUNMathStat : Mathematical Statistics for Functional Data Analysis

Summary of the project :

The interest for functional data analysis is growing in various fields of application. The statistical problems posed are very specific and original because they concern infinite dimensional objects which require a dedicated mathematical framework, at the interface between statistics of processes and functional analysis. Our project is structured in three axes which correspond to major statistical problematics such as dimension reduction, regression, and testing. The challenge is to contribute to unlocking the theoretical issues specific to functional data and to establish inference procedures based on strong theoretical results and to develop estimation and tests procedure useful for practionners. We wish to extend existing dimension reduction tools of functional data or regression methods to more general models allowing to handle more complex data. Moreover, the subject of testing procedures have been little addressed so far and we wish to develop dedicated tools supported by strong theoretical results. 

Therefore, our consortium gathers specialists in mathematical statistics and applied statistics around an original project, with the ambition to build a network of young and more experienced collaborators on this theme.

We organize the FDA&A seminar dedicated to Functional Data Analysis and its Applications. Click here to subscribe to the mailing list.