ECML PKDD 2024

The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD) is a flagship European event for the Machine Learning and Data Mining community which will take place between September 9th and 13th in Vilnius, Lithuania. The conference features peer-reviewed research presentations, talks from industry leaders, workshops, tutorials, poster sessions, and various tracks including applied data science, industry, and PhD forums. Lamarr scientists are organizing four workshops, a keynote, and two tutorials (as part of the workshops) at the event, along with contributing with numerous publications.

Workshop Interactive Adaptive Learning (IAL) (September 9th)

The “Interactive Adaptive Learning (IAL)” full day-workshop focuses on optimizing the entire machine learning process, particularly in scenarios with scarce or evolving data, by integrating techniques like adaptive, active, semi-supervised, and transfer learning. It aims to bring together researchers and practitioners to explore novel approaches, practical experiences, and unresolved challenges in developing interactive and adaptive machine learning systems. It is co-organized by Lamarr researchers Dr. Mirko Bunse and Dr. Amal Saadallah (both TU Dortmund University).

Workshop TempXAI (September 9th)

The full-day workshop at ECML PKDD 2024, titled “TempXAI: Explainable AI for Time Series and Data Streams Tutorial-Workshop” aims to address the intersection of explainable AI (XAI) and the challenges of interpreting time series and data streams. Key objectives include: understanding how to interpret Machine Learning models in evolving data environments; ensuring model interpretability remains effective as data changes over time; and applying XAI to real-time contexts for interpretable and timely decision-making.

Lamarr’s researchers Matthias Jakobs and Dr. Amal Saadallah (both TU Dortmund University), Lamarr’s Fellow Prof. Dr. Barbara Hammer (Bielefeld University) and Lamarr’s Principal Investigator for Trustworthy AI, Prof. Dr. Emmanuel Müller, are all contributing into the organization of the workshop.

Workshop Learning to Quantify: Methods and Applications (LQ 2024) (September 13th)

The “Learning to Quantify: Methods and Applications” workshop at ECML/PKDD 2024 explores the task of training models to estimate class prevalence in unlabelled data, a process distinct from traditional classification. This event aims to introduce researchers to the principles, methods, and applications of Learning to Quantify, while also fostering discussion on advancements and practical applications in handling large-scale data. One of the organizators is Dr. Mirko Bunse from Lamarr’s partner organization, TU Dortmund University.

Workshop Quantum Machine Learning (September 13th)

The workshop and tutorial “Quantum Machine Learning” (QML) focuses on the latest developments in QML, emphasizing the balance between innovation and theoretical rigor. Alternatives to Quantum Neural Networks (QNNs) and hybrid methods, the strengths and limitations of QML, along with theoretical contributions will all be discussed by participants in interactive sessions.

Lamarr’s researchers Dr. Nico Piatkowski (Fraunhofer IAIS), Sascha Mücke (TU Dortmund University) and Prof. Dr. Christian Bauckhage, Lamarr’s director at the Fraunhofer IAIS site are all contributing into the organization of the workshop.

Keynote “Resource-Aware Machine Learning — a User-Oriented Approach” (September 12th)

Prof. Dr. Katharina Morik, foundating director of the Lamarr Institute will also be a part of the ECML PKDD. Morik’s keynote addresses the growing need for energy-efficient machine learning (ML) by exploring research on making models smaller, faster, and more resource-conscious through techniques like pruning, quantization, and energy-efficient hardware. It emphasizes the importance of user-friendly testing tools to help stakeholders choose the most energy-efficient models, advocating for greater consideration of user roles in developing sustainable ML solutions.

Access the full program of the conference here. And register here for the conference.