DeSBi Retreat 2025
8 - 9 July 2025
DeSBi Retreat 2024
7 - 8 July 2024
Our research unit, RU KI-FOR 5363, recently held its annual retreat—an event we eagerly anticipate each year. This retreat serves as a critical platform for promoting scientific exchange and taking full advantage of the diverse expertise within our project teams. More than just a gathering, it’s a space where we come together to discuss the progress of our work, showcase key achievements, and engage in meaningful discussions that drive innovation forward.
DAGStat 2025
24 - 28 March 2025
The DeSBi Research Unit organized a dedicated session on its research topics at DAGStat 2025 in Berlin. The session, titled "Fusing Deep Learning and Statistics Towards Understanding Structured Biomedical Data," was chaired by our associated postdoctoral researcher, Georg Keilbar. Oral presentations were delivered by unit PhD students and postdocs: Masoumeh Javanbakht, Marco Simancher, Manuel Pfeuffer, and Sepideh Saran.
The presented topics covered a diverse range of cutting-edge research, including deep nonparametric conditional independence tests for images, visual explanations for statistical tests, deep modeling in the presence of known confounders with applications to neuroimaging data, and an empirical analysis of uncertainty quantification in genomics applications.
DeSBi Joint Seminar Series
KI-FOR 5363 DeSBi
Fusing Deep Learning and Statistics towards Understanding Structured Biomedical Data (DeSBi)
Publications
Paulo Yanez, Simon Witzke, Nadja Klein, Bernhard Y. Renard: Sparse Explanations of Neural Networks Using Pruned Layer-Wise Relevance Propagation
Published in: Machine Learning and Knowledge Discovery in Databases. Research Track. ECML PKDD 2024. Lecture Notes in Computer Science
Abstract:
Explainability is a key component in many applications involving deep neural networks (DNNs). However, current explanation methods for DNNs commonly leave it to the human observer to distinguish relevant explanations from spurious noise. This is not feasible anymore when going from easily human-accessible data such as images to more complex data such as genome...
Marco Simnacher, Xiangnan Xu, Hani Park, Christoph Lippert, Sonja Greven: Deep Nonparametric Conditional Independence Tests for Images
Published in: arXiv
Abstract:
Conditional independence tests (CITs) test for conditional dependence between random variables. As existing CITs are limited in their applicability to complex, high-dimensional variables such as images, we introduce deep nonparametric CITs (DNCITs). The DNCITs combine embedding maps, which extract feature representations of high-dimensional variables, with nonparametric CITs applicable to these feature representations.







