Prof. Dr. Ulf Brefeld

21335 Lüneburg, Universitätsallee 1, C4.311
Fon +49.4131.677-1663, ulf.brefeld@leuphana.de
Professor, Professorship of Information Systems, in particular Machine Learning

Table of contents for this page

Vita

I am a professor for Machine Learning at Leuphana Universität Lüneburg. Prior to joining Leuphana, I was a joint professor for Knowledge Mining & Assessment at TU Darmstadt and the German Institute for Educational Research (DIPF), Frankfurt am Main. Before, I led the Recommender Systems group at Zalando SE and worked on machine learning at Universität Bonn, Yahoo! Research Barcelona, Technische Universität Berlin, Max Planck Institute for Computer Science in Saarbrücken, and at Humboldt-Universität zu Berlin. I received a Diploma in Computer Science in 2003 from Technische Universität Berlin and a Ph.D. (Dr. rer. nat.) in 2008 from Humboldt-Universität zu Berlin. I am interested in statistical machine learning and data mining. ML3 homepage

Projects

Publications

Books and anthologies

  1. Machine Learning and Data Mining for Sports Analytics: 11th International Workshop, MLSA 2024, Vilnius, Lithuania, September 9, 2024, Revised Selected Papers
    Ulf Brefeld (Editor) , Jesse Davis (Editor) , Jan Van Haaren (Editor) , Albrecht Zimmermann (Editor) , 2025 Cham , 119 p.

    Research output: Books and anthologies › Conference proceedings › Research

  2. Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2019, Würzburg, Germany, September 16–20, 2019, Proceedings, Part I
    Ulf Brefeld (Editor) , Elisa Fromont (Editor) , Andreas Hotho (Editor) , Arno Knobbe (Editor) , Marloes Maathuis (Editor) , Céline Robardet (Editor) , 2020 Cham , 766 p.

    Research output: Books and anthologies › Conference proceedings › Research

  3. Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2019, Würzburg, Germany, September 16–20, 2019, Proceedings, Part II
    Ulf Brefeld (Editor) , Elisa Fromont (Editor) , Andreas Hotho (Editor) , Arno Knobbe (Editor) , Marloes Maathuis (Editor) , Céline Robardet (Editor) , 2020 Cham , 732 p.

    Research output: Books and anthologies › Conference proceedings › Research

  4. Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2019, Würzburg, Germany, September 16–20, 2019, Proceedings, Part III
    Ulf Brefeld (Editor) , Elisa Fromont (Editor) , Andreas Hotho (Editor) , Arno Knobbe (Editor) , Marloes Maathuis (Editor) , Céline Robardet (Editor) , 2020 Cham , 804 p.

    Research output: Books and anthologies › Conference proceedings › Research

  5. Machine Learning and Data Mining for Sports Analytics: 7th International Workshop, MLSA 2020, Co-located with ECML/PKDD 2020, Ghent, Belgium, September 14–18, 2020, Proceedings
    Ulf Brefeld (Editor) , Jesse Davis (Editor) , Jan van Haaren (Editor) , Albrecht Zimmermann (Editor) , 2020 Cham , 146 p.

    Research output: Books and anthologies › Conference proceedings › Research

View all publications

Journal contributions

  1. Principled Transformers for Predictive Performance in Knowledge Tracing
    Kai Neubauer (Author) , Ulf Brefeld (Author) , Yannick Rudolph (Author) , 01.01.2026 , in: Journal of Educational Data Mining, 18, 1 , p. 89-112 , 24 p.

    Research output: Journal contributions › Journal articles › Research › peer-review

  2. Joint Item Response Models for Manual and Automatic Scores on Open-Ended Test Items
    Ulf Brefeld (Author) , Daniel Bengs (Author) , Ulf Kroehne (Author) , Fabian Zehner (Author) , 01.09.2025 , in: Psychometrika, 90, 4 , p. 1346-1367 , 22 p.

    Research output: Journal contributions › Journal articles › Research › peer-review

  3. The promise and challenges of computer mouse trajectories in DMHIs – A feasibility study on pre-treatment dropout predictions
    Jennifer Matthiesen (Author) , Ulf Brefeld (Author) , Burkhardt Funk (Author) , Kirsten Zantvoort (Author) , Pontus Bjurner (Author) , Marie Bendix (Author) , Viktor Kaldo (Author) , 01.06.2025 , in: Internet Interventions, 40 , 7 p.

    Research output: Journal contributions › Journal articles › Research › peer-review

  4. Masked autoencoder for multiagent trajectories
    Ulf Brefeld (Author) , Yannick Rudolph (Author) , 01.02.2025 , in: Machine Learning, 114, 2 , 18 p.

    Research output: Journal contributions › Journal articles › Research › peer-review

  5. Interactive sequential generative models for team sports
    Ulf Brefeld (Author) , Moritz Cordes (Author) , Dennis Fassmeyer (Author) , 01.02.2025 , in: Machine Learning, 114, 2 , 15 p.

    Research output: Journal contributions › Journal articles › Research › peer-review

View all publications

Contributions to collected editions/anthologies

  1. Self-improvement for Computerized Adaptive Testing
    Kai Neubauer (Author) , Ulf Brefeld (Author) , Yannick Rudolph (Author) , 01.01.2026 Cham , p. 70-86 , 17 p.

    Research output: Contributions to collected editions/anthologies › Article in conference proceedings › Research › peer-review

  2. Masked Autoencoder Pretraining for Event Classification in Elite Soccer
    Ulf Brefeld (Author) , Yannick Rudolph (Author) , 26.02.2024 Cham , p. 24-35 , 12 p.

    Research output: Contributions to collected editions/anthologies › Article in conference proceedings › Research › peer-review

  3. Hands in Focus: Sign Language Recognition Via Top-Down Attention
    Ulf Brefeld (Author) , Noha Sarhan (Author) , Christian Wilms (Author) , Vanessa Closius (Author) , Simone Frintrop (Author) , 08.10.2023 Piscataway , p. 2555-2559 , 5 p.

    Research output: Contributions to collected editions/anthologies › Article in conference proceedings › Research › peer-review

  4. Exploring the Poincaré Ellipsis
    Tino Paulsen (Author) , Ulf Brefeld (Author) , Samuel Fadel (Author) , 01.09.2023

    Research output: Contributions to collected editions/anthologies › Article in conference proceedings › Research › peer-review

  5. User Authentication via Multifaceted Mouse Movements and Outlier Exposure
    Jennifer J. Matthiesen (Author) , Ulf Brefeld (Author) , Hanne Hastedt (Author) , 01.04.2023 Cham , p. 300-313 , 14 p.

    Research output: Contributions to collected editions/anthologies › Article in conference proceedings › Research

View all publications

Activities

  1. Self-improvement for Computerized Adaptive Testing
    Kai Neubauer (Speaker) , Ulf Brefeld (Coauthor) , Yannick Rudolph (Speaker)

    Activity: Conference Presentations › Research

  2. Masked Autoencoder Pretraining for Event Classification in Elite Soccer
    Ulf Brefeld (Coauthor) , Yannick Rudolph (Speaker)

    Activity: talk or presentation in privat or public events › Research

  3. Data-efficient Pattern Detection in Elite Soccer
    Dennis Faßmeyer (Coauthor) , Ulf Brefeld (Coauthor) , Yannick Rudolph (Speaker)

    Activity: Conference Presentations › Research

  4. Modeling Conditional Dependencies in Multiagent Trajectories
    Ulf Brefeld (Coauthor) , Yannick Rudolph (Speaker)

    Activity: Presentations (poster etc.) › Research

  5. Detection of tactical patterns using semi-supervised graph neural networks
    Gabriel Anzer (Speaker) , Pascal Bauer (Speaker) , Ulf Brefeld (Speaker) , Dennis Faßmeyer (Speaker)

    Activity: Conference Presentations › Research

View all activities

Press / Media

  1. Data Science: Lern- und Ausbildungsinhalte
    1 Media contribution

    Press/Media: Press/Media

View all press/media coverage

Courses

Deep Learning
Ulf Brefeld, Maximilian Hannes Reinhardt
Next appointment:
Tuesday, 2026-04-21 at 12:15
Room: C 9.102 Seminarraum
Deep Learning
Ulf Brefeld, Maximilian Hannes Reinhardt
Next appointment:
Tuesday, 2026-04-28 at 10:15
Room: C 14.102 a Seminarraum
Elevator Pitch Seminar
Ulf Brefeld
⌄
The seminar addresses presentation and writing in technical domains. We'll begin with elevator pitches, that is, very short summary presentations. The name comes from the possibility of meeting a VIP in front of an elevator and you have only a very limited amount of time before the VIP exists the elevator to tell her about your idea/research/undertaking/... We will then turn the oral presentation into a scientific write-up. The audience will take the role of independent reviewers commenting on motivation/technical contribution/risk/challenges/... etc.

In general, the shorter the presentation the more difficult it is and the more work you should put into it to give a good presentation. In this seminar, we will get to know the ingredients of a good presentation, how to prepare for it and how to perform well in front of an audience (e.g., VIP in an elevator).
Next appointment:
Lectures for this semester ended.
Forecasting and Simulation
Ulf Brefeld, Soham Majumder
⌄
The module provides a survey of the theory and application of data-based computational techniques to forecast and simulate data with temporal dependencies. Selected statistical approaches dealing with the special role of time in modeling will be discussed in detail. Topics of interest include:
- stationary and non-stationary time series (ARIMA models)
- conditional heteroscedastic time series (ARCH and GARCH models)
- multivariate time series (VAR and VARMA models)
- state space models (Kalman Filter)
Next appointment:
Thursday, 2026-04-23 at 10:15
Room: C 14.102 a Seminarraum
Forecasting and Simulation
Ulf Brefeld, Soham Majumder
⌄
The module provides a survey of the theory and application of data-based computational techniques to forecast and simulate data with temporal dependencies. Selected statistical approaches dealing with the special role of time in modeling will be discussed in detail. Topics of interest include:
- stationary and non-stationary time series (ARIMA models)
- conditional heteroscedastic time series (ARCH and GARCH models)
- multivariate time series (VAR and VARMA models)
- state space models (Kalman Filter)
Next appointment:
Tuesday, 2026-04-21 at 14:15
Room: C 14.001 Seminarraum
Introduction to Artificial Intelligence
Ulf Brefeld
⌄
Es werden grundlegende Kenntnisse des maschinellen Lernen vermittelt, die zur intelligenten Datenverarbeitung notwendig sind. Wir gehen auf verschiedene Problemstellungen ein (z.B. Klassifikation, Regression, Clustern), lernen Standardverfahren kennen (z.B. Entscheidungsbäume, SVMs), Strategien für eine allgemeine Herangehensweise und Versuchsaufbauten für eine solide empirische Evaluation.
Next appointment:
Tuesday, 2026-04-21 at 12:30
Room: C HS 3
Introduction to Artificial Intelligence - Exerceise Group 1
Ulf Brefeld, Tino Paulsen
⌄
Es werden grundlegende Kenntnisse des maschinellen Lernen vermittelt, die zur intelligenten Datenverarbeitung notwendig sind. Wir gehen auf verschiedene Problemstellungen ein (z.B. Klassifikation, Regression, Clustern), lernen Standardverfahren kennen (z.B. Entscheidungsbäume, SVMs), Strategien für eine allgemeine Herangehensweise und Versuchsaufbauten für eine solide empirische Evaluation.
Next appointment:
Tuesday, 2026-04-28 at 08:15
Room: C 5.019 Seminarraum
Introduction to Artificial Intelligence - Exerceise Group 3
Ulf Brefeld, Shirley Metz
⌄
Es werden grundlegende Kenntnisse des maschinellen Lernen vermittelt, die zur intelligenten Datenverarbeitung notwendig sind. Wir gehen auf verschiedene Problemstellungen ein (z.B. Klassifikation, Regression, Clustern), lernen Standardverfahren kennen (z.B. Entscheidungsbäume, SVMs), Strategien für eine allgemeine Herangehensweise und Versuchsaufbauten für eine solide empirische Evaluation.
Next appointment:
Tuesday, 2026-04-28 at 10:15
Room: C 12.112 Seminarraum
Introduction to Artificial Intelligence - Exerceise Group 4
Ulf Brefeld, Tino Paulsen
⌄
Es werden grundlegende Kenntnisse des maschinellen Lernen vermittelt, die zur intelligenten Datenverarbeitung notwendig sind. Wir gehen auf verschiedene Problemstellungen ein (z.B. Klassifikation, Regression, Clustern), lernen Standardverfahren kennen (z.B. Entscheidungsbäume, SVMs), Strategien für eine allgemeine Herangehensweise und Versuchsaufbauten für eine solide empirische Evaluation.
Next appointment:
Tuesday, 2026-04-28 at 10:15
Room: C 16.109 /110 Seminarraum
Introduction to Artificial Intelligence - Exercise Group 2
Ulf Brefeld, Shirley Metz
⌄
Es werden grundlegende Kenntnisse des maschinellen Lernen vermittelt, die zur intelligenten Datenverarbeitung notwendig sind. Wir gehen auf verschiedene Problemstellungen ein (z.B. Klassifikation, Regression, Clustern), lernen Standardverfahren kennen (z.B. Entscheidungsbäume, SVMs), Strategien für eine allgemeine Herangehensweise und Versuchsaufbauten für eine solide empirische Evaluation.
Next appointment:
Tuesday, 2026-04-28 at 08:15
Room: C 16.223 Seminarraum
On Stupidity
Ulf Brefeld, Claus Pias
⌄
»Intelligence is what the tests test.« (Edwin G. Boring: »Intelligence as the Tests Test It,« in: New Republic 36 (1923), S. 35-37)

Das Seminar beschäftigt sich mit informatischen und kulturwissenschaftlichen Aspekten von Künstlicher Intelligenz. Damit diese nicht getrennt bleiben, wollen über die Disziplinen hinweg ins Gespräch kommen.

Dabei gehen wir von drei Schwerpunkten aus:

1. Nicht nur die sogenannte »KI«, sondern alle Intelligenz ist »künstlich«, weil sie immer eine kulturelle (d.h. nicht rein ›menschliche‹ oder angeborene) Leistung ist. Obwohl das Wort alt ist, sind die Techniken ihrer Beobachtung jung: etwa in Form von Intelligenztests, die »Intelligenz« daran messen, wie gut Versuchspersonen Intelligenztest lösen können. Sobald ein IQ gemessen ist, gibt es nicht nur Normale und Kluge, sondern plötzlich auch Dumme und noch Dümmere – über die längste Zeit Juden, Nicht-Europäer, Kriminelle, Frauen etc. (Stephen Jay Gould). Wir werden also der Frage nachgehen, wie Dummheit und Naturalisierung von »Intelligenz« zusammenhängen.

2. Es gibt »dumme« Arbeiten. Die Computerisierung sollte sie, wie der Kybernetiker Norbert Wiener in den 1950ern prognostizierte, an Maschinen als moderne »Sklaven« delegieren, damit Menschen für intelligentere Tätigkeiten entlastet würden. Heute würde man z.B. fragen: Ist Auswendiglernen dumm und das Schreiben von Hausarbeiten ein Aufweis von Intelligenz? Und seit wann eigentlich? Wir werden daher der Frage nachgehen, was »dumme« Arbeiten waren und sind, in deren Bereich heute künstliche Intelligenz interveniert.

3. Die Zeichen selbst sind zwar »dumm« (Jacques Lacan), aber wenn man sie lange genug geschickt kombiniert und nach Wahrscheinlichkeiten aufeinander folgen läßt (Shannon), sieht das im Ergebnis nach »Intelligenz« aus. Joseph Weizenbaum ärgerte sich in den 1960ern über die Dummheit von Usern, die seinen recht einfachen Chatbot ELIZA für intelligent hielten. Und über die gerade gegründete Informatik, dass sie das ausnutzt. Andere KI-Pioniere waren weniger erfolgreich und haben mit viel mehr Aufwand nur Ergebnisse erzielt, die von außen eher dumm wirkten. Wir werden daher der Frage nachgehen, was und warum und aus wessen Perspektive eigentlich dummes oder intelligentes Verhalten ist.
Next appointment:
Wednesday, 2026-04-22 at 12:15
Room: C 12.006 Seminarraum
Research Colloquium Business Informatics & Data Science
Olga Abramova, Ulf Brefeld, Paul Drews, Burkhardt Funk, Peter Niemeyer, Kathrin Padberg-Gehle, Ricardo Usbeck
⌄
The Research Colloquium at the Institute for Information Systems (IIS) at Leuphana University is a dynamic and interdisciplinary platform for scholarly exchange, critical dialogue, and academic networking focused on PhD students in and around Data Science and AI. Designed to foster a vibrant research culture, the colloquium brings together faculty, researchers, doctoral candidates, and external guests to discuss current topics in the field of information systems and beyond.

Our colloquium features a diverse set of formats:

* Research Talks: IIS faculty and doctoral researchers present ongoing projects and recent findings, receiving constructive feedback in a collaborative environment.

* Invited Expert Lectures: Renowned national and international scholars are invited to share cutting-edge research and theoretical insights, offering new perspectives and fostering academic exchange.

* Startup & Industry Talks: Entrepreneurs and professionals from startups and established companies discuss real-world challenges and innovations at the intersection of business and technology, providing practical insights and networking opportunities.

* Interdisciplinary Dialogues: The colloquium also opens space for collaborative sessions across disciplines, reflecting the integrative and transdisciplinary mission of Leuphana University.

Open to members of the university and interested guests, the IIS Research Colloquium serves as an essential forum for inspiration, collaboration, and advancing impactful research.
Next appointment:
Wednesday, 2026-04-22 at 14:00
Room: C 7.019 Seminarraum