How can machines learn from interaction? Or how can intelligence, or more in general, complex behavior emerge from simple parts?

Big questions ofcourse, but nonetheless fascinating, and I enjoy studying these questions as a researcher in the Data Mining Group as well as in the Media Technology Program at the Leiden Institute of Advanced Computer Science at Leiden University in the Netherlands.

Next to that I apply these methods in industry, most recently as the Director of Decisioning Solutions at Pegasystems, and I also have experience with start ups, both as an employee and an informal advisor.

I am still fond of classical machine learning and data mining research, but also more and more intrigued by what I like to call 'Artificial X', artificializing phenomena beyond intelligence, from creativity to ethics. Not just to understand more, but also to better appreciate what we don't yet understand. And not just to understand what's possible tomorrow, but to speculate on deep future.


As introduced in my welcome statement, my research focusses on questions such as 'How can machines learn from interaction?' or 'Or how can intelligence, or more in general, complex behavior emerge from simple parts?'. This translates into research in data mining and machine learning, as well as in broader creative research.

A resulting example of the combination of these two areas can be seen in the image above. This was a MSc Media Technology project in computational creativity by Jules Verdijk. A system was created that evolves abstract paintings, that maximize a specific emotion. We asked people to indicate the emotion associated with paintings from a set of real world abstract paintings, and then classifiers were trained to predict this emotion from lower level image features. These classifiers were then used to evaluate the fitness of newly evolved paintings.

For details, see my list of scientific publications and an overview of current and past supervised MSc and Phd thesis research projects. I also write or speak about these topics for general public audiences.


I primarily teach in the Media Technology Program at Leiden University. The Media Technology MSc programme is a place where students are encouraged to formulate their own scientific questions, and to translate personal inspirations and curiosities into their own research projects. To answer these questions, students create actual products, because by doing and creating, new scientific insights into the underlying question are encountered. I teach the New Media New Technology class, coach semester project teams and supervise MSc and Phd thesis research projects. Occasionally, I also give ad hoc or guest lectures or workshops on machine learning, robots or creative research. In 2021 I will start teaching a course on Artificial Creatures.

Check out Teaching for more detail.

Other academic activities

I have been on program committees for conferences such as KDD, CIKM, EMCL PKDD, BNAIC and Benelearn, was sponsor chair for PKDD and organized workshops at KDD and PKDD, and organized Benelearn 2011 (conference chair) and am Society and Industry Chair for BNAIC-Benelearn 2020. I have also organized data mining competitions based on The Insurance Company (TIC) Benchmark, which has also been contributed to the UCI KDD repository. I also regularly write and present for the general public.

Check out Other Activities for more detail.

Thanks for your visit

Check the top right menu for publications, teaching resources and more. If you are working in related areas of research and are interested in collaboration in the areas of data mining or creative research, do not hesitate to get in touch.