Wouter Duivesteijn

Hi, I'm Wouter, and I'm incurably curious. I am an Assistant Professor in Data Mining at the Technische Universiteit Eindhoven. My research revolves around Exceptional Model Mining (EMM): a local pattern mining method where we seek subsets of the dataset that are interesting, which they are if they satisfy two conditions. On the one hand, they must be interpretable: we must be able to succinctly describe the definition of a subgroup, so that the knowledge that they represent becomes actionable. On the other hand, they must be exceptional: they must display some kind of behavior that sets them apart from the overall population. The scientific challenges revolve around how to efficiently search for subgroups, and how to express exceptional behavior such that the subgroups we find are meaningful.

CV (last updated: March 27, 2023)

Now Hiring!

I have an open position for a fully funded 5-year PhD TA position on the topic of Efficient Algorithms for Exceptional Model Mining on Time-Varying Data. You can read more about the research project in this document.

How to Cheat the Page Limit: the 2022 Update

I am one of the Proceedings Chairs of ECML PKDD 2022. In this role, we wrote a report on our findings, updating our paper on How to Cheat the Page Limit.

Sound of Science #16

Flemish comedian and science fanatic Lieven Scheire presents a podcast at TU/e, called Sound of Science. In this podcast, he discusses the last scientific discoveries and the role of technology in society with researchers and students. In Episode 16, he and I talked about my work (in Dutch). NLP researchers are invited to use this episode as a training set to let their algorithms distinguish a Vlaams from a Rotterdams accent in the Dutch language.

Latest publications

You can find my full publication list sorted by category here, and sorted by year here.

Gelfilter dataset

In addition to a new data mining method, our IDA 2020 paper also introduces a dataset, publicly available for research purposes free of charge. It is a run to failure time series library, where the primary task is to predict the Remaining Useful Life of a gel filter in a chemical plant. You can find more details and the dataset itself here.
Contact information