Functional data are more and more often appearing
in many fields of applied statistics (medicine, environmetrics, chemometrics,
finance,
bioinformatics,..), and there is really necessity for developing statistical
tools for their treatment. Indeed, most of usual statistical problems well-known
for real or multivariate data, have their natural counterpart in the functional
setting. Let us just mention for instance the following problems:
-> curves classification;
-> curves discrimination;
-> regression from functional explanatory and response variables
-> conditional functional quantiles
-> ...
There is real challenging problems, both from methodological and applied points
of view, in developing functional adaptation of usual
techniques to these new kinds of problems. Among the previous advances in this
field, we mention:
-> linear and generalized linear models for functional data;
-> nonparametric modelling for functional regression
-> bootstrapping functional data
-> learning with functional data (boosting, bagging,neural networks....)
-> testing hypothesis and model selection;
-> ....
This track will cover these new statistical topics in functional data analysis.
The working group STAPH on Functional and Operatorial
Statistics in Toulouse, France and the group on nonparametric statistics in the
University of Santiago de Compostela will be scientifically associated with the
organization of this session.