E-mail: r.m.schouten@tue.nl
Please contact me for my research and teaching statements.
Methods: Exceptional Model Mining | Local Pattern Mining | Subgroup Discovery | Subgroup Analysis |
Type of Data: Tabular | Non-IID | Hierarchical | Sequential | Time Series | Longitudinal | Multilevel |
Application domains: Personalized Learning Analytics | e-Health | Medicine |
Related keywords: Heterogeneous Treatment Effects, Counterfactual Explanations | Dynamic Bayesian Networks | Causal Diagrams | Mixed-Effects Models | Missing Data |
In my research, I develop data mining methods that discover and describe differences between individuals. People differ, and my methods aim to reveal exceptional, coherent and interpretable subgroups in a population of persons. For instance, in the medical context, we aim to discover subgroups of patients with exceptional responses to treatments. And in the educational domain, we discover subgroups of students with exceptional learning behaviour.
My methodological contributions stand out because I develop generic methods that have real-world impact. I enjoy working at the intersection of Data Mining, Statistics and Social Sciences.
I obtained a personal NWO take-off grant, and work together with Turku Research Institute for Learning Analytics and AlgebraKit to investigate whether our data mining technology can be implemented in digital learning platforms.
Together with Erasmus Medical Center and Catharina Hospital Eindhoven, I find solutions for population heterogeneity and missing data in longitudinal data. In the past, I worked with Netherlands Institute of Mental Health and Addiction, Hospital Twente and Utrecht Medical Center.
Before joining TU/e, I worked as a Missing Data Researcher and visited Prof. Andrew Gelman at the Department of Statistics at Columbia University in 2016. Recently, I visited Prof. Barbara Hammer at Bielefeld University in 2025. Furthermore, in 2024, I served as Proceedings Chair at ECML PKDD, and was pronounced an Excellent Reviewer.
I love interacting with students, and have supervised over 20 individual students, 15 groups of students, taught in 4 Master level courses, and received Two awards for Excellent Course Evaluations. I currently supervise 2 PhD candidates.
2025 - present: Post-doc at TU/e.
2020 - 2024: PhD candidate at TU/e.
Supervisors: Prof. Mykola Pechenizkiy, Dr. Wouter Duivesteijn.
Topic: Exceptional Model Mining for Hierarchical Data.
2017 - 2019: Missing Data Researcher at Utrecht University.
Supervisors: Prof. Stef van Buuren, Dr. Gerko Vink.
2016: Staff Associate of Professor Andrew Gelman at Columbia University in the City of New York, US. Duration: 2 months.
2015: Intern at SRON, Dutch Institute of Space Research, the Netherlands. Duration: 4 months.
Recognition as Excellent reviewer Research Track ECML PKDD 2024
Award for Excellent course evaluation Research Topics in Data Mining 2022/2023
Award for Excellent course evaluation Research Topics in Data Mining 2021/2022
In top-3 Best Poster during NWO Commit2Data Day in 2021
| Year | Call | Type | Title | Together with |
|---|---|---|---|---|
| 2024 | Take-off Phase 1 NWO | €40k Feasibility study | Integration of Local Pattern Mining in Digital Assessment Tools | Prof. Mykola Pechenizkiy |
| 2022 | AI for Health EWUU Alliance | EUR 45k seed money | Better Imputation by Generative Adversarial NeTworks (BIGANT) | Prof. Stef van Buuren, Dr. Gerko Vink, Hanne Oberman, Prof. Mykola Pechenizkiy, Prof. Cassio de Campos, Rianne Schouten, Daniel Oberski, Dr. Thomas Debray, Prof. Fred van Eeuwijk |
| Year | Where? | Who? | Why? | Duration |
|---|---|---|---|---|
| 2025 | Leuven University | Prof. Hendrik Blockeel, Prof. Jesse Davis | To discuss Pattern mining and Personalized recommendations | 3 days, in August |
| 2025 | Ghent University | Prof. Tijl de Bie, Prof. Jefrey Lijfijt | To discuss Pattern mining and fairness | 2 days, in August |
| 2025 | Bielefeld University | Prof. Barbara Hammer | To discuss Concept drift, XAI and User interaction | 1 week |
| 2025 | University of Twente | Prof. Monique Tabak | To discuss E-health and Personalized recommendations | 1 day |
| 2024 | Radboud University Nijmegen | Dr. Marcos L.P. Bueno | To provide a guest-lecture in the AI for Healthcare course | 1 day |
| 2016 | Columbia University | Prof. Andrew Gelman | To collaborate on Evaluating missing data methods, together with Dr. Gerko Vink | 2 months |
See Google Scholar: >20 publications, >400 citations and h-index: 7.
van den Biggelaar, L., Schouten, R.M., de Bie, A., Bouwman, A, & Duivesteijn. W. (2025) Characterizing the Risk of Atrial Fibrillation in Cardiac Patients with Exceptional Electrocardiogram Phenotypes. Accepted for publication at KDD25.
Schouten, R.M. (2025) Exceptional Model Mining for Hierarchical Data. PhD thesis.
Schouten, R.M. (2024) On the role of prognostic factors and effect modifiers in structural causal models. Accepted for presentation at Causal Representation Learning Workshop NeurIPS. See here for the paper!
Van den Berg, N. T., Broekgaarden, B. O., Mahieu Dionysia, P., Martens, J. G., Niederle, J., Schouten, R.M., & Duivesteijn, W. (2024) Generating MNAR missingness in image data, with additional evaluation ofMisGAN. Accepted for presentation at BNAIC/BeNeLearn 2024. See here for the presentation.
Schouten, R.M., Stevens, G.W.J.M., van Dorsselaer, S.A.F.M., Duinhof, E.L., Monshouwer, K., Pechenizkiy, M. & Duivesteijn, W. (2024) Analyzing the interplay between societal trends and socio-demographic variables with local pattern mining: Discovering exceptional trends in adolescent alcohol use in the Netherlands. To appear in post-proceedings BNAIC/BeNeLearn 2024.
Schouten, R.M., Duivesteijn, W., Rasanen, P, Paul, J.M., & Pechenizkiy, M. (2024) Exceptional Subitizing Patterns: Exploring Mathematical Abilities of Finnish Primary School Children with Piecewise Linear Regression. In: Proceedings of the European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), pp. 66-82.
Schouten, R.M., Tascau, V., Ziegler, G.G., Casano, D., Ardizonne, M., & Erotokritou M.A. (2023) Dropping incomplete records is (not so) straightforward. In: Proceedings of the 21st International Symposium on Intelligent Data Analysis (IDA), pp. 379-391.
Verhaegh, R.F.A., Kiezebrink, J.J.E., Nusteling, F., Rio, A.W.A, Bendicsek, M.B., Duivesteijn, W. & Schouten, R.M. (2022) A Clustering-inspired Quality Measure for Exceptional Preferences Mining — Design Choices and Consequences. In: Proceedings of the International Conference on Discovery Science (DS), pp. 429–444.
Schouten, R.M., Zamanzadeh, D. & Singh, P. (2022) Pyampute: a Python library for data amputation. Zenodo. https://doi.org/10.25080/majora-212e5952-03e.
Van der Haar, J.F., Nagelkerken, S.C., Smit, I.G., van Straaten, K., Tack, J.A., Schouten, R.M. & Duivesteijn, W. (2022) Efficient Subgroup Discovery Through Auto-Encoding In: Proceedings of the 20th International Symposium on Intelligent Data Analysis (IDA), pp. 327-340.
Schouten, R.M., Duivesteijn, W. & Pechenizkiy, M. (2022) Exceptional Model Mining for Repeated Cross-Sectional Data (EMM-RCS). In: Proceedings of SIAM International Conference on Data Mining (SDM), pp. 585-593.
Schouten, R.M., Bueno, M.L.P., Duivesteijn, W. & Pechenizkiy, M. (2022) Mining Sequences with Exceptional Transition Behaviour of Varying Order using Quality Measures based on Information-Theoretic Scoring Functions Data Mining and Knowledge Discovery (DAMI), 36: 379-413.
IJsselhof R, Duchateau S, Schouten R.M., Slieker M, Hazekamp M & Schoof P. (2020) Long-Term Follow-Up of Pericardium for the Ventricular Component in Atrioventricular Septal Defect Repair World Journal for Pediatric and Congenital Heart Surgery, 11(6): 742-747.
IJsselhof R.J., Duchateau S.D.R., Schouten R.M., Freund, M.W., Heuser, J., Fejzic, Z., Haas, F., Schoof, P.H. & Slieker, M.G. (2019) Follow-up After Biventricular Repair of the Hypoplastic Left Heart Complex European Journal of Cardiothoracic Surgery, 57(4): 644-651.
Schouten R.M., Lugtig, P. & Vink, G. (2018) Generating missing values for simulation purposes: A multivariate amputation procedure Journal of Statistical Computation and Simulation, 88(15): 1909-1930.
Schouten, R.M. and Vink, G. (2021) The dance of the mechanisms: How observed information influences the validity of missingness assumptions Sociological Methods & Research, 50(3): 1243-1258.
Kappen, I.F.P.M., Bittermann, G.K.P., Schouten, R.M., Bittermann, D., Etty, E., Koole, R., Kon, M., Van der Molen, M. & Breugem, C.C. (2017) Long-term mid-facial growth of patients with a unilateral complete cleft of lip, alveolus and palate treated by two-stage palatoplasty: cephalometric analysis Clinical Oral Investigations, 21: 1801-1810.
de Vries, C.P., Schouten, R.M., Van der Kuur, J., Gottardi, L., & Akamatsu, H. (2016) Microcalorimeter pulse analysis by means of principle component decomposition In: Proceedings SPIE 9905, Space TElescopes and Instrumentation 2016: Ultraviolet to Gamma Ray, 99055v. DOI: 10.1117/12.2231627
| Year | Organization | Type of collaboration |
|---|---|---|
| 2025-now | AlgebraKit | Answering domain questions using EMM |
| 2022-now | Turku Research Institute for Learning Analytics | Answering domain questions using EMM |
| 2022-now | Dutch south west Psoriatic Arthritis Registry, Erasmus MC | Solving MD problems + Answering domain questions using EMM |
| 2020-now | Biomedical Systems and Signals Group UTwente and Hospital Twente | Discovering exceptional blood glucose fluctuations |
| 2021-2024 | Netherlands Institute of Mental Health and Addiction | Answering domain questions using EMM |
| 2017-2019 | University Medical Center Utrecht | Statistics consultant |
At Eindhoven University of Technology, my teaching track record is:
| Year | Course | Level | Activities |
|---|---|---|---|
| 24/25 | Research Topics in Data Mining | MSc | Responsible for track: Empirical Challenges in Data Mining |
| 22/23 | Research Topics in Data Mining | MSc | Taught 2 lectures about Missing Data, supervised 4 groups of students during research project |
| 21/22 | Research Topics in Data Mining | MSc | Taught 2 lectures about Missing Data, supervised 7 groups of students during research project |
| 20/21 | Foundations of Data Mining | MSc | Taught 2 lectures about Missing Data, provided answers in weekly Q&A, developed practice questions for exam, administrative activities |
“I took a lot of courses last year, but I like your instructions the most. It is not only because of your professional knowledge, but also because of your personality of being kind, patient, responsible.” (Jin Ouyang, Master student, 2022)
At Utrecht University, my teaching track record is:
| Year | Course | Level | Activities |
|---|---|---|---|
| 2016 | Survey Research: Statistical Analysis and Estimation | Advanced MSc | Organization,Tutoring exercise classes |
| 2016 | Survey Research: Design, Implementation and Data Processing | Advanced MSc | Organization,Tutoring exercise classes |
| 2015 | Survey Research: Statistical Analysis and Estimation | Advanced MSc | Organization,Tutoring exercise classes |
| 2015 | Survey Research: Design, Implementation and Data Processing | Advanced MSc | Organization,Tutoring exercise classes |
“Rianne was a first class assistant at our summer school courses. Not only was all material prepared extremely punctual and without errors, she also got very high student evaluations. I can wholeheartedly recommend Rianne!” (Prof. dr. Edith de Leeuw, 2016)
Current students:
| Start | Student | Type | Topic |
|---|---|---|---|
| May 2025 | Abdullahi Farah | MSc Thesis | Numerical Optimization for EMM |
| Aug 2024 | Lieke van den Biggelaar | PhD Research | Exceptional Model Mining with Time Series Data |
| Nov 2023 | Emmanuel C. Chukwu | PhD Research | Counterfactual Explanations in Time Series Classification |
Alumni:
| Year | Student | Type | Topic |
|---|---|---|---|
| 24/25 | Haoqi Guo | MSc Thesis | Improving Diversity and Feasibility of Counterfactual Explanations, supervision together with prof.dr. Mykola Pechenizkiy |
| 23/24 | Lieke van den Biggelaar | MSc Thesis with Catharina Hospital | Discovering subgroups of patients with exceptional Atrium Fibrillation based on ECGs, supervision together with dr. Wouter Duivesteijn |
| 22/23 | Victoria Tascau | MSc Thesis with DEPAR/Erasmus Medical Centrum | Handling Missing Values in Longitudinal Medical Data, supervision together with dr. Wouter Duivesteijn |
| 21/22 | Mika van Loon | MSc Thesis | Bootstrap Hypothesis Tests for Evaluating Subgroup Descriptions in Exceptional Model Mining, supervision together with dr. Wouter Duivesteijn |
| 21/22 | Varun Kamat | Internship at Signify | Recommendation Tool for Component Database |
| 21/22 | Isabel van den Heuvel | Research Proposal | Equivalence Testing for Developing Fair Machine Learning Algorithms, supervision together with Hilde Weerts |
| 21/22 | Mats Verbraak | Research Proposal | Handling Missing Data in the Prediction Domain using Multiple Imputation |
| 20/21 | Bart van Dooren | MSc Thesis with Philips | Predicting Cardiovascular Risk with Objective Physical Activity Measurements, supervision together with prof.dr. Mykola Pechenizkiy |
I reviewed >15 papers for top-level data mining conferences and statistical journals (DAMI, ECML PKDD, EWAF, JRSSB, SiM, BimJ). I was recognized as an Excellent reviewer at ECML PKDD 2024. I served as Proceedings Chair at ECML PKDD 2024. At EWAF 2025, I was a Session chair.
ampute in R-package miceR-function ampute is the implementation of a
multivariate amputation procedure: a method for generating missing data
in complete datasets. With ampute, it is straightforward to
generate missing values in multiple variables, with different missing
data proportions and varying underlying missingness mechanisms. Read the article or the
vignette to learn more.
parlMICEFor large datasets or when you want to impute with a large number of
imputations, multiple imputation with mice in R-package
mice may have a long run time. As a solution, Gerko
Vink and I created wrapper function parlMICE, which allows
for a parallel run of mice.
The function is now part of package mice under the
name parlmice.
All information can be found in the github repo or in the vignette.
pyampute: the first Python library for data
amputationLibrary pyampute provides the multivariate amputation
methodology to the Python community, and it does more. It has improved
default settings, allows for a combined MAR+MNAR mechanism, for custom
probability functions and since it is compatible with scikit-learn’s
fit and transform paradigm, seamless
integration in data processing pipelines becomes easy.
Find plenty of examples in pyampute’s
documentation. Davina’s presentation at SciPy22 can be found here.
Install using pip or from source:
pip install pyamputegit clone https://github.com/RianneSchouten/pyampute.git
pip install ./pyampute| Year | Occasion | Type | Topic | Link to materials |
|---|---|---|---|---|
| 2025 | Research visit @ Bielefeld University | Colloquium talk | Exceptional Model Mining for Hierarchical Data | |
| 2025 | Info-topic about my PhD research | News | Exceptional Model Mining for Hierarchical Data | link |
| 2024 | AI for Healthcare | Invited guest-lecture at Radboud University Nijmegen | Exceptional Model Mining | |
| 2021 | EAISI Eindhoven | Invited presentation | Towards a better understanding of exceptional lifestyle behaviour | |
| 2019 | Workshop R-Ladies Amsterdam | Invited presentation | Developed and presented a workshop about analysis of missing values, evaluation and implementation of missing data methods | link |
| 2018 | ICT Open | Contribution to 1-day conference | Handling Missing Data in Data Science | link |
| 2018 | European Women in Technology | Masterclass at conference | Dealing with missing data in R: Amputation or
Imputation? |
presentation and exercises |
| 2018 | sat-R-Day | Contribution to 1-day conference | Missing data | link |
| 2018 | Data Science Hackathon | By invitation | Developed and lead a missing data challenge | link |
| 2017 | Amst-R-Dam | By invitations | How to use R-function ampute to generate missing values
in complete datasets |
article and documentation |
| 2017 | UseR!2017 | Contribution to conference | Introduction to multivariate amputation with
ampute |