Matthias Jakobs
Topics
- Explainable Machine Learning with a focus on the application of Shapley values
- Time series forecasting, especially under concept drift
- Online model selection and ensembling
Selected Publications
-
Harnessing Prior Knowledge for Explainable Machine Learning: An Overview (SaTML, 2023)
-
Shapley Values with Uncertain Value Functions (IDA, 2023)
-
Explainable Online Ensemble of Deep Neural Network Pruning for Time Series Forecasting (Machine Learning, 2022)
-
Explainable Online Deep Neural Network Selection Using Adaptive Saliency Maps for Time Series Forecasting (ECML-PKDD, 2021)
Social Media
GitHub | Google Scholar | Mastodon | Twitter/X
Other Scientific Activities
- Program Committee member, Workshop on Trustworthy Artificial Intelligence, ECML-PKDD 2022
- Co-organizer, XAI-TS Workshop, ECML-PKDD 2023
Supervised Theses
- Lisa Salewsky, 2021 (B.Sc.): Interaktive Visualisierung verschiedener Erklärbarkeitsverfahren für Neuronale Netze
- Rahel Wilking, 2021 (M.Sc.): Untersuchung der Anfälligkeit perturbationsbasierter Erklärbarkeitsmethoden für Adversarial Attacks
- Hanna Mykula, 2021 (B.Sc.): Online Adaptive Multivariate Time Series Forecasting