Jilles Vreeken

Interpretable machine learning

Causality in machine learning

Unsupervised learning

Jilles Vreeken conducts research at the CISPA Helmholtz Center for Information Security on methods for deriving actionable and potentially novel causal insights from large and complex data sets and machine learning models. His work enables researchers not only to explore their data in new ways and generate new hypotheses, but also to understand how data-driven models arrive at their conclusions.

A second major focus of his research is causality. To date, most AI systems base their predictions primarily on correlations, that is, on statistical regularities in the data. Even highly powerful models, however, are generally unable to distinguish between correlation and causation or to infer causal relationships. Vreeken aims to bridge this gap by developing theoretical foundations and methodological approaches for causal discovery and reasoning in machine learning.

Vreeken earned his doctorate in computer science from Utrecht University in 2009. Following research positions at the University of Antwerp, Saarland University, and the Max Planck Institute for Informatics, he joined CISPA in 2018. There, he leads the Exploratory Data Analysis research group. In addition, Vreeken is an honorary professor of computer science at Saarland University and a Fellow of the European Laboratory for Learning and Intelligent Systems (ELLIS).