The Lamarr Institute has expanded its faculty with two exceptional researchers whose expertise will significantly advance our mission of developing cutting-edge, trustworthy, and efficient AI methods. This winter semester, Dr. Katharina Eggensperger and Jun. Prof. Dr. Matthias Feurer have joined our team to further deepen Lamarr’s research in Automated Machine Learning (AutoML) and its role in creating reliable and accessible machine learning solutions.
Advancing Accessible and Efficient ML: Prof. Dr. Katharina Eggensperger
Katharina Eggensperger joins the Lamarr Institute from Eberhard Karls University of Tübingen, where she led the Early Career Research Group “AutoML for Science” at the Cluster of Excellence Machine Learning for Science. Her work aligns closely with Lamarr’s goal to enable scientifically grounded, transparent, and efficient AI that empowers researchers across disciplines.
Her research focuses on making machine learning both more accessible and more robust, with a strong emphasis on machine learning for tabular data and scientific applications. Dr. Eggensperger’s contributions to open-source tools, hyperparameter optimization, and ML benchmark systems have shaped the AutoML field internationally. She is also a driving force within the (Auto)ML community through major roles in workshops, conferences, and the AutoML School.
Motivated by the goal to provide easy access to efficient state-of-the-art ML, I research how to improve and extend both AutoML methods and ML models, to leverage the full potential of ML for applications in science. In this context, I currently aim to better understand, systematically evaluate, design and apply foundation models for tabular data.
Katharina Eggensperger
With her arrival, Lamarr strengthens its mission to create ML methods that are both high-performing and usable in real-world scientific environments, ensuring reliable and reproducible results.
Advancing Trustworthy and Scalable AI: Jun. Prof. Dr. Matthias Feurer
Matthias Feurer joins the Lamarr Institute from Ludwig-Maximilians-Universität München, where he served as a Thomas Bayes Fellow of the Munich Center for Machine Learning and interim professor in the department of statistics. His research directly supports Lamarr’s guiding principles of trustworthiness and performance in AI development.
His work in AutoML, meta-learning, and optimization simplifies the development of ML systems without compromising on fairness, interpretability, or reliability. Matthias Feurer’s contributions to major open-source projects such as Auto-sklearn, OpenML, and SMAC3 exemplify Lamarr’s commitment to transparent and globally accessible AI tools.
With AutoML, I’m removing the complexity from AI development so machine learning experts and domain experts can focus on solving problems and building trustworthy and accurate solutions that drive innovation. A central aspect of my work is empirical research to generate reliable insights and guide methodological progress.
Matthias Feurer
Through his research, Jun. Prof. Feurer enhances Lamarr’s mission to build AI technologies that are both powerful and responsible, lowering barriers to adoption while ensuring transparency and user empowerment.
Together, the appointments of Katharina Eggensperger and Matthias Feurer mark an important milestone in strengthening Lamarr’s position as a leading center for scientific excellence in AI methodology. Their work will contribute to the institute’s vision of developing dependable AI solutions that advance science and benefit society.