Abstract:

Training any machine learning pipeline always comes with a multitude of different design decisions, incl. hyperparameters, neural architectures, machine learning algorithms, preprocessing, and even more. Choosing them manually can be tedious and error-prone. Automated machine learning (AutoML) supports developers and researchers by proposing solutions to all these design decisions. Therefore, AutoML can help find ML solutions with better predictive performance, faster inference time, smaller memory footprint, improved fairness, or smaller CO2 footprint. Although automating the entire ML design might seem appealing and contribute to the democratization of AI, AutoML cannot be successful without any human being involved. In my talk, I will cover the main concepts of AutoML and argue that a more human-centric approach is needed not only in AI but also in AutoML.

Short Bio:

Prof. Dr. Marius Lindauer is full professor of machine learning at Leibniz University Hannover. He received his PhD from the University of Potsdam (Germany) in 2015 under Prof. Dr. Thorsten Schaub and Prof. Dr. Holger Hoos. From 2014 to 2019, he was PostDoc and later junior research group lead at the University of Freiburg under Prof. Dr. Frank Hutter. Besides being a member of ELLIS, he is one of the co-heads of automl.org and co-founder of the research network COSEAL, the AutoML conference and the Institute of AI (LUH|AI) at the Leibniz University Hannover. He won several competitions, including SAT solving, ASP solving, and automated machine learning, and his open-source packages are downloaded more than 70k times each month. He gave several tutorials at renowned AI conferences and summer schools, such as AAAI, IJCAI and ESSAI. In 2022, he was awarded an ERC starting grant, the most prestigious research grant for European young researchers. His research interests include automated machine learning (AutoML), reinforcement learning, interpretable and sustainable AI.