University of Groningen

The Groningen node, the main node in SUNDIAL, is also the largest, with 4 PhD students. Unique is that it contains both a group of astronomers/astrophysicists and a group of computer scientists. The astronomers are part of the Kapteyn Astronomical Institute, and the computer scientists part of the Johann Bernoulli Institute, both research institutes in the Faculty of Mathematics and Natural Sciences at the University of Groningen. Both institutes collaborate formally through the Faculty Theme DSSC (Data Science and Science Complexity).

The Kapteyn Astronomical Institute, responsible for the bachelor and master Astronomy, is part of the Dutch top-research school NOVA, a collaboration with the other 3 astronomy-institutes in the Netherlands. It is a department with a broad selection of topics in astronomy, with 16 FTE scientific staff, about 50 PhD students and 20 postdocs. The department is in the same building as SRON, the Space Research Organisation of the Netherlands, and at 60 km from the Radio Astronomy institute ASTRON. The institute has a strong data science group, including the OmegaCEN group, responsible for the ground segment of Euclid, and has 2 instrumentation groups, one in Groningen building submm instrumentation, and one in Dwingeloo building optical/NIR instrumentation for, e.g., the E-ELT.

The PI of the SUNDIAL project is the astronomer Reynier Peletier. He is mainly interested in studying the formation and evolution of galaxies, by studying their structure, kinematics and stellar populations. At the moment, his research is mostly focused on studying dwarf galaxies in the nearby Universe, as will be the case for his role in the SUNDIAL project. He is, however, also very much interested in larger galaxies and galaxy cluster formation. He is co-PI (together with Enrica Iodice in Naples) of the Fornax Deep Survey FDS, one of the 2 surveys around which SUNDIAL has been designed. Peletier will be the supervisor of one ESR project. He can be contacted for all SUNDIAL-related issues:

Other Kapteyn members involved in SUNDIAL are Scott Trager, Edwin Valentijn, Gijs Verdoes Kleijn, Leon Koopmans and Peter Barthel, as well as PhD students Aku Venhola, Elaheh Hamraz and Seyda Sen, and the SUNDIAL manager (tbd.).

The computer scientists of SUNDIAL are all part of the intelligent systems group of the Johann Bernoulli Institute. The group works on a wide range of topics of machine learning and mathematical morphology. There will be 3 ESRs in this group, one led by each of the supervisors Michael Biehl, Kerstin Bunte and Michael Wilkinson.

With a background in statistical and computational physics, Michael Biehl's research has always been centered on interdisciplinary areas such as the theory and application of neural networks or the simulation of complex dynamical systems. Many scientific disciplines are experiencing an impressive increase of the rate at which data are acquired. Several methodological challenges can be identified immediately: High-dimensional, complex data sets need to be made accessible by means of compression and visualization techniques. Problems related to clustering, classification, or regression trigger the search for efficient practical schemes for the selection of relevant features in the data. The integration of information from various sources poses the question of how to combine and relate heterogeneous data sets efficiently. These challenges are also in the focus of the current research interests. Recent activities are centered on computational modeling in general and, more specifically, the machine learning based analysis of complex, high-dimensional data sets. Three major, inter-dependent aspects of this work can be identified: Theoretical studies, algorithm development, and real world applications. The objective of theoretical investigations and simulation of machine learning processes is to achieve a thorough understanding of phenomena and problems, which can occur in practical situations.

Information about previous and on-going research, links to publications, talks and presentations, software tools etc. are available at Michael's homepage. For further information please contact him at

With a background in Computer Science for Natural Sciences Kerstin Bunte has from the start focused on interdisciplinary research. Besides a spectrum of machine learning developments for classification, clustering, dimensionality reduction and visualization her recent Marie Curie Individual Fellowship centered at machine learning involving complex dynamic systems in the biomedical domain. Therefore, the project of ESR12 ideally complements recent developments and the experience in machine learning of the intelligent systems group and its members via an very interesting collaboration between the university and companies with respect to astronomical questions.

Information about Kerstin is available from her homepage. She can be contacted at

Michael Wilkinson is an expert in digital image analysis. His expertise is in fields, such as segmentation, mathematical morphology and connected filters. He applies his knowledge to biomedical imaging and visualization, and astronomical image analysis, for example of the detection of the faintest objects. He is also an active amateur astronomer.

Information about Michael is available from his homepage. He can be contacted at