Jan Schuchardt
Technical University of Munich
Department of Informatics - I26
Boltzmannstr. 3
85748 Garching b. München
Germany
Room: 00.11.061
Phone: +49 (0)89 / 289-17380
E-Mail: j.schuchardt [at] tum.de
GitHub: jan-schuchardt
Research Focus
My research is focused on making machine learning models for graphs and other structured data more trustworthy. Specifically, I am interested in methods that provably guarantee robustness and privacy in domains with equivariances.
Publications
(Provable) Adversarial Robustness for Group Equivariant Tasks: Graphs, Point Clouds, Molecules, and More
Jan Schuchardt, Yan Scholten, Stephan Günnemann
Conference on Neural Information Processing Systems (NeurIPS), 2023
[PDF]
Hierarchical Randomized Smoothing
Yan Scholten, Jan Schuchardt, Aleksandar Bojchevski, Stephan Günnemann
Conference on Neural Information Processing Systems (NeurIPS), 2023
[PDF]
Localized Randomized Smoothing for Collective Robustness Certification
(selected for spotlight presentation)
Jan Schuchardt*, Tom Wollschläger*, Aleksandar Bojchevski, Stephan Günnemann
International Conference on Learning Representations (ICLR), 2023
[PDF]
Invariance-Aware Randomized Smoothing Certificates
Jan Schuchardt, Stephan Günnemann
Conference on Neural Information Processing Systems (NeurIPS), 2022
[PDF]
Randomized Message-Interception Smoothing: Gray-box Certificates for Graph Neural Networks
Yan Scholten, Jan Schuchardt, Simon Geisler, Aleksandar Bojchevski, Stephan Günnemann
Conference on Neural Information Processing Systems (NeurIPS), 2022
[PDF]
Generalization of Neural Combinatorial Solvers Through the Lens of Adversarial Robustness
Simon Geisler & Johanna Sommer, Jan Schuchardt, Aleksandar Bojchevski, Stephan Günnemann
International Conference on Learning Representations (ICLR), 2022
[PDF]
Collective Robustness Certificates: Exploiting Interdependence in Graph Neural Networks
Jan Schuchardt, Aleksandar Bojchevski, Johannes Gasteiger, Stephan Günnemann
International Conference on Learning Representations (ICLR), 2021
[PDF]
Workshop papers
Training Differentially Private Graph Neural Networks with Random Walk Sampling
Morgane Ayle, Jan Schuchardt, Lukas Gosch, Daniel Zügner, Stephan Günnemann
Workshop on Trustworthy and Socially Responsible Machine Learning
Conference on Neural Information Processing Systems (NeurIPS), 2022
[PDF]
Research Background
- 2020: Master's thesis: Collective Robustness Certificates
- 2018: Bachelor's thesis: Reinforcement Learning for Adaptation in Evolutionary Computation
Education
- 2018 - 2020: M.Sc. Computer Science, Technical University of Munich
- 2015 - 2018: B.Sc. Computer Science, Technical University of Munich