Cookie usage policy

The website of the University Carlos III of Madrid use its own cookies and third-party cookies to improve our services by analyzing their browsing habits. By continuing navigation, we understand that it accepts our cookie policy. "Usage rules"

Paul H.C. Eilers


Paul H.C.Eilers

Paul H.C.Eilers
Erasmus University Medical Centre   NETHERLANDS

Paul Eilers (1948) was trained, long ago, as an electronic engineer, but he has worked in a variety of fields: measurement and analysis of mechanical vibrations, aquatic ecology of the Dutch delta, environmental protection for the region around the Rotterdam harbor, medical statistics, social science and genetics and bioinformatics. The binding theme is his fascination for the statistical analysis of complex data.
His favorite subject is smoothing and he has done a lot of work in the development and application of spline-based techniques. Their use in higher dimensions demands advanced statistical computations. The development of efficient algorithms is an attractive challenge to him.
He likes to work together with statisticians and other scientists, and he has done a lot of consulting. This has resulted in over 150 papers in peer-reviewed journals.
He has a long-time cooperation with Maria Durbán at Carlos III University, and others, on smoothing methods for large multi-dimensional data sets.


Project: Ill-posed functional data analysis of indirect observations.The project combines functional data analysis, indirect observations and ill-posed problems.
Functional data analysis recognizes the fact that many data sets represent continuous data, such as time series or spatial fields, in a discrete way. It can be fruitful to model the situation using mathematical functions of one or more variables.
Functional data are seldom observed directly. The measurement technique can pose limitations, like a PET scan in brain imaging. Or geographical units do not allow fine detail. Or data might be grouped on purpose, to protect privacy. A statistical model relates observation to latent functions to be estimated.
Often data and model are not informative enough to get a reliable estimate of the latent function and the problem is called ill-posed. A solution is to bring in prior information and combine it in a proper way, using penalties, with the data.

Stay period: OCT 14 - JUN 15