Wouter Duivesteijn
Wouter Duivesteijn is a former member of the ADA group. He is currently a post-doctoral researcher at the TU Dortmund.
Wouter was a Ph.D. student in the ADA group from September 2009 to June 2013. For years Wouter claimed to be a mathematician who also happens to do computer science. After receiving his Master's degrees with a final thesis on a data mining subject, he finally converted to computer science in 2009, when he became a Ph.D. student in the Algorithms group of LIACS. Each week he spend two days in the ADA group. Wouter worked on the NWO Exceptional Model Mining (EMM) project. EMM is a framework that can be seen as a generalisation of Subgroup Discovery (SD). Both SD and EMM attempt to find small portions of the data where the observed behaviour is notably different from that of the database as a whole. But, whereas in SD `behaviour' is traditionally interpreted in terms of the distribution of a single nominal variable, EMM seeks subgroups for which the fitted local model is surprisingly different from the global model. In this approach, `behaviour' is described by a number of attributes, and fitting a model captures the multivariate dependencies between these attributes.

Selected Refereed Publications

Duivesteijn, W., Feelders, A., Knobbe, A. Different Slopes for Different Folks - Mining for Exceptional Regression Models with Cook's Distance. In: Proceedings KDD 2012, 2012.
Duivesteijn, W., Knobbe, A.J., Feelders, A. & van Leeuwen, M. Subgroup Discovery meets Bayesian networks – an Exceptional Model Mining approach. In: Proceedings of the 10th IEEE International Conference on Data Mining (ICDM'10), 2010.
Duivesteijn, W. & Feelders, A. Nearest Neighbour Classification with Monotonicity Constraints. In: Proceedings of the European Conference on Machine Learning and Principles and Practices of Knowledge Discovery in Data 2008 Part I (ECML PKDD'08), pp 301-316, 2008.