Since storing data electronically is steadily becoming easier and cheaper, one is nowadays often collecting nearly continuous data records. Examples can be found in diverse fields, including environmental sciences (e.g. pollution curves), geophysics (e.g. earth surface temperature), medicine (e.g. fMRI images), econometrics (e.g. tick-data), and many others. To benefit from increasing information, we need appropriate statistical tools that can help extracting the most important characteristics of some possibly high-dimensional specifications. However, most of the existing tools are not suitable for this type of data, which are often sampled sequentially in time (e.g. day-by-day sequenced data). Ignoring the resulting dependence may yield suboptimal and sometimes even spurious or inappropriate conclusions.

This research project is devoted to the study of serially correlated functional data. Researchers from the Mathematics Department will try to develop a broadly applicable theoretical framework for the data generating processes and develop new statistical tools for such functional time series data.


HÖRMANN Siegfried
Mathematics Department
Faculty of Sciences

Created on August 31, 2018