The seamless integration of large data sets into sophisticated computational models provides one of the central research challenges for the mathematical sciences in the 21st century. When the computational model is based on evolutionary equations and the data set is time-ordered, the process of combining models and data is called data assimilation. The assimilation of data into computational models serves a wide spectrum of purposes ranging from model calibration and model comparison all the way to the validation of novel model design principles.
The field of data assimilation has been largely driven by practitioners from meteorology, hydrology and oil reservoir exploration; but a theoretical foundation of the field is largely missing. Furthermore, many new applications are emerging from, for example, biology, medicine, and the neurosciences, which require novel data assimilation techniques. The goal of the proposed CRC is therefore twofold: First, to develop principled methodologies for data assimilation and, second, to demonstrate computational effectiveness and robustness through their implementation for established and novel data assimilation application areas.
While most current data assimilation algorithms are derived and analyzed from a Bayesian perspective, the CRC will view data assimilation from a general statistical inference perspective. Major challenges arise from the high-dimensionality of the inference problems, nonlinearity of the models and/or non-Gaussian statistics. Targeted application areas include the geoscience as well as emerging fields for data assimilation such as biophysics and cognitive neuroscience.
Prof. Dr. Sebastian Reich, University of Potsdam, Department of Mathematics
Dr. Liv Heinecke, University of Potsdam, Department of Mathematics
Our next CoderDojo will take place on Saturday the 30.11. from 11 am to 3 pm. Can you help our little robots to save Christmas? A whole bunch of new...
We co-organized the 3. Career Day that took place on the 7th of November with the overarching headline “Research-based job opportunities – new careers...
Recent progress in machine learning has shown how to forecast and, to some extent, learn the dynamics of a model from observations, resorting in...more ›
Slow-fast systems consist of slow macroscopic and fast microscopic dynamics. By using equation-free methods, one can do a complete bifurcation...more ›
We discuss methods for creating Gaussian likelihoods for data that does not directly follow any known statistics. Obvious summary statistics are...more ›
Gugushvili, S., Mariucci, E. and Meulen, van der F. (2019). Decompounding discrete distributions: A non-parametric Bayesian approach. To appear in Scandinavian Journal of Statistics. arXiv: 1903.11142
Mariucci, E., Ray, K. and Szabó, B. (2019). A Bayesian nonparametric approach to log-concave density estimation. To appear in Bernoulli. arXiv: 1703.09531
Roelly, S. and Zass, A. (2019). Marked Gibbs point processes with unbounded interaction: an existence result. arXiv: 1911.12800