About me
I am an Early Career Research Group Leader AutoML for Science in the Cluster of Excellence Machine Learning for Science at the University of Tübingen.
Short Bio
My research focuses on methods for automated machine learning (AutoML), hyperparameter optimization (HPO), and efficient benchmarking. Motivated by the goal of further democratizing the application of machine learning for scientific researchers and practitioners, I research how to improve and extend AutoML systems to leverage the full potential of ML for new applications.
Previously, I was part of the ML Lab at the University of Freiburg, where I completed my Ph.D. under the supervision of Frank Hutter and Marius Lindauer (2022). I also co-developed several open-source tools for HPO methods and AutoML systems and have been a member of the team winning three AutoML competitions (2016, 2018, 2020). I am a junior head of the Automl.org group, member of the COSEAL group and a faculty member of IMPRS-IS. Furthermore, I co-organized the AutoML workshop series at ICML in 2019, 2020 and 2021, served as a social chair (2022,2023) and as a program chair (2024) for the AutoML Conference and am co-leading the Tübingen Women in Machine Learning group.
News
- Oct'24 · 📓 I will offer a seminar on "(Auto-)ML for tabular data" in the winter semester. See here.
- Sep'24 · 🗨️ I will give an invited Talk at the "AutoML Conference X Paris WiMLDS"
- July'24 · 🗨️ I gave an invited talk at the "ML in Science Conference" in Tübingen on "Automated Machine Learning for Science"
- June'24 · 👋 Welcome our new group member: Mykahilo Koshil!
- May'24 · 📝 Our paper "Position: Why We Must Rethink Empirical Research in Machine Learning" has been accepted at ICML'24.
- May'24 · 🗨️ I gave an invited talk at Women in Data Science Regensburg on "AutoML: Streamlining Machine Learning".
- Mar'24 · 📝 Our paper "Towards Bandit-based Optimization for Automated Machine Learning" has been accepted for oral presentation at PML4LRS@ICLR'24.
- Mar'24 · 📝 Our paper "Towards quantifying the effect of dataset selection for benchmarking tabular machine learning approaches" has been accepted at DMLR@ICLR'24.
- Feb'24 · 📝 Our paper "Can Fairness be Automated? Guidelines and Opportunities for Fairness-aware AutoML" has been published at JAIR.
- Feb'24 · 📓 I will offer a seminar on "AutoML in the Age of Large Pre-trained Models" in the summer semester. See here.
- Sep'23 · 🥳 I am thrilled to be a program chair for AutoML 2024 in Paris (together with Roman Garnett and Joaquin Vanschoren , and Marius Lindauer as the General Chair)
more entries
- 9th Nov'23 · 🗨️ Together with Marius Lindauer I gave a talk at MLOps'23: "Hyperparameter Optimieren mit AutoML"
- 15th Sep'23 · 👋 Welcome my first group member: Amir Rezaei Balef!
- Sep'23 · 📓 I offer a seminar on "Automated Machine Learning and Hyperparameter Optimization" in the winter semester. Find all details here.
- 12th-15th Sept'23 · 🥳 Together with Matthias Feurer I am organizing the social program for AutoML 2023! Find all details here.
- July'23 · 🗨️ Together with Marius Lindauer I will give a lecture on "AutoML: Accelerating Research on and Development of AI Applications" at the ESSAI Summer School!
- June'23 · 🗨️ Together with Eddie Bergman I gave a *Hands-On Session* at the nextgen_AI Freiburg workshop: "Automated Machine Learning with Auto-sklearn"
- Jan'23 · 📝 Our paper Mind the Gap: Measuring Generalization Performance Across Multiple Objectives got accepted at IDA 2023.
- Jan'23 · 📝 Our paper TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second got accepted to ICLR 2023.
- Jan'23 · 🥳 I am very excited to start my own Early Career Research Group for AutoML for Science at the Cluster of Excellence "ML for Science" at the University of Tübingen!
Misc
Note: For links to slides, tutorials and videos, please see talks, teaching and Automl.org
Want to work with me? I typically recruit PhD students via IMPRS-IS (Deadline mid-November each year). Here is an example call for a PhD position or reach out to me for more details!