Matthias Herp

Statistics and Bio-Informatics PhD student
In the programme Data Science

Research
I work in a collaboration between the chair of statistics with Prof. Dr. Thomas Kneib and the department of medical bioinformatics with Prof. Dr. Michael Altenbuchinger . I focus on developing nonparametric graphical methods to model complex dependency structures in high dimensional data using transformation models. Our goal is to provide methods for the analysis of omics data in bioinformatics.
Keywords: Multivariate Statistics, Graphical Models, Transformation Models

Publications

• Graphical Transformation Models (2025) , joint work with Johannes Brachem , Prof. Dr. Michael Altenbuchinger and Prof. Dr. Thomas Kneib (currently an Arxiv Preprint & under Review)
  • Python package: Github Repository

Conferences & Workshops

• YES workshop | Young European Statisticians | Edition XIII (YES XIII | Dependence Modeling 2025)
• High-dimensional statistics, applications, and distributional shifts. ( A workshop in celebration of Peter Bühlmann’s 60th birthday)
• 7th Joint Statistical Meeting of the Deutsche Arbeitsgemeinschaft Statistik (DAGStat 2025): presentation of Graphical Transformation Models (2025)
• 26th International Conference on Computational Statistics (COMPSTAT 2024): presentation of the Graphical Transformation Model

Teaching

• Exercise Class Tutor for the Masters Course: Mathematical Foundations of Applied Statistics [M.MED.0010], WiSe 2025/26
• External lecturer for the Generalised Additive Models (GAMs) Modul 2 of the Medical Data Science Certificate, at the Heidelberg University Hospital, 2025

Thesis Topics

• Extending the Graphical Conditional Transformation Model (a short description can be found here)