Explanatory Data Analysis group

Dr. Matthijs van Leeuwen

Dr. Matthijs van Leeuwen
Dr. Matthijs van Leeuwen
Associate professor & group leader

Associate professor & group leader Website Google Scholar profile LinkedIn profile Twitter profile

The short (compressed) version

Matthijs likes data, patterns, algorithms, and information theory. He strives for data mining and machine learning methods and results that are principled, interpretable, and exploit existing knowledge.

The longer version

Matthijs is associate professor and group leader at the Leiden Institute of Advanced Computer Science (LIACS), Leiden University. He is Programme Manager of the Master Computer Science, and affiliated with SAILS and DSRP, the university-wide research programmes for artificial intelligence (AI) and data science. His primary research interests are exploratory data mining and interpretable machine learning: how can we enable domain experts to explore and analyse their data, to uncover patterns and make predictions, and—ultimately—discover novel knowledge?

For this it is important that methods and results are explainable to domain experts, who may not be data scientists. His signature approach is to define and identify patterns that matter, i.e., succinct descriptions that characterise relevant structure present in the data. Which patterns matter strongly depends on the data and task at hand, hence defining the problem is one of the key challenges in his research. Information theoretic concepts such as the Minimum Description Length (MDL) principle have proven very useful to this end. Matthijs is also interested in interactive data mining, i.e., involving humans in the loop. Finally, he is interested in fundamental data mining research for real-world applications, both in science (e.g., life sciences, social sciences) and industry (e.g., manufacturing and engineering, aviation), as this is the best way to show that the theory works in practice.

Bio

Matthijs was previously a (tenure track) assistant professor (2017-2020) and senior researcher (2015-2017) at Leiden University, and a postdoctoral researcher at KU Leuven (2011-2015) and Universiteit Utrecht (2009-2011). He defended his Ph.D. thesis, titled Patterns that Matter, in February 2010, at Universiteit Utrecht. He won several best paper and reviewer awards at international conferences and was awarded NWO Rubicon, FWO Postdoc, NWO TOP2, and NWO TTW Perspectief grants. He is General Chair of the IDA Council and editorial board member of Data Mining and Knowledge Discovery. Further, he co-organised a number of international conferences and workshops, and co-lectured tutorials on 'Information Theoretic Methods in Data Mining'.

More information, including CV, at www.patternsthatmatter.org

Selected recent publications

In press
Yang, L, Baratchi, M & van Leeuwen, M Unsupervised Discretization by Two-dimensional MDL-based Histogram. Machine Learning, Springerwebsite
2023
Lopez-Martinez-Carrasco, A, Proença, HM, Juarez, JM, van Leeuwen, M & Campos, M Discovering Diverse Top-k Characteristic Lists. In: Proceedings of the 21st International Symposium on Intelligent Data Analysis (IDA 2023), Springer, 2023.
Papagianni, I & van Leeuwen, M Discovering Rule Lists with Preferred Variables. In: Proceedings of the 21st International Symposium on Intelligent Data Analysis (IDA 2023), Springer, 2023.
Kroes, SKS, van Leeuwen, M, Groenwold, RHH & Janssen, MP Generating synthetic mixed discrete-continuous health records with mixed sum-product networks. Journal of the American Medical Informatics Association vol.30(1), Oxford University Press, 2023.
2022
Li, Z & van Leeuwen, M Feature Selection for Fault Detection and Prediction based on Event Log Analysis. ACM SIGKDD Explorations vol.24(2), ACM, 2022.
Spaink, HA, Verhagen, IE, van Leeuwen, M & Terwindt, GM Methodological considerations in predicting migraine attacks using machine learning. In: MTIS 2022 Cephalalgia Abstracts, Sage Publications, 2022.
Li, Z & van Leeuwen, M Feature Selection for Fault Detection and Prediction based on Log Analysis. In: Proceedings of the international workshop on AI for Manufacturing Workshop at ECMLPKDD 2022, 2022.
Yang, L, Opdam, T & van Leeuwen, M Histogram-based Probabilistic Rule Lists for Numeric Targets. In: Proceedings of the 20th anniversary Workshop on Knowledge Discovery in Inductive Databases (KDID 2022) at ECMLPKDD 2022, CEUR Workshop Proceedings, 2022.
Yang, L & van Leeuwen, M Truly Unordered Probabilistic Rule Sets for Multi-class Classification. In: Proceedings of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECMLPKDD 2022), Springer, 2022.implementationwebsite
Vinkenoog, M, van Leeuwen, M & Janssen, M Explainable hemoglobin deferral predictions using machine learning models: interpretation and consequences for the blood supply. Vox Sanguinis
Proença, HM, Grünwald, P, Bäck, T & van Leeuwen, M Robust subgroup discovery - Discovering subgroup lists using MDL. Data Mining and Knowledge Discoveryimplementationwebsite
van Rijn, S, Schmitt, S, van Leeuwen, M & Bäck, T Finding Efficient Trade-offs in Multi-Fidelity Response Surface Modeling. Engineering Optimizationwebsite
Yang, L & van Leeuwen, M Probabilistic Rule Sets Ready for Interactive Machine Learning. In: AAAI'22-Workshop on Interactive Machine Learning, 2022.
Vinkenoog, M, Steenhuis, M, ten Brinke, A, van Hasselt, C, Janssen, M, van Leeuwen, M, Swaneveld, F, Vrielink, H, van de Watering, L, Quee, F, van cen Hurk, K, Rispens, T, Hogema, B & van der Schoot, E Associations between symptoms, donor characteristics and IgG antibody response in 2082 COVID-19 convalescent plasma donors. Frontiers in Immunology, Frontiers