Research Interests

• Nonparametric and high-dimensional statistics
• Statistical learning
• Statistics for stochastics processes

Preprints

• The surrogate Gibbs-posterior of a corrected stochastic MALA: Towards uncertainty quantification for neural networks with Sebastian Bieringer, Gregor Kasieczka and Mathias Trabs, arXiv:2310.09335. A previous version of the preprint was entitled "Statistical guarantees for stochastic Metropolis-Hastings".
• AdamMCMC: Combining Metropolis adjusted Langevin with momentum-based optimization with Sebastian Bieringer, Gregor Kasieczka and Mathias Trabs, arXiv:2312.14027
• Estimating a multivariate Lévy density based on discrete observations, arXiv:2305.14315

Publications

Preprint versions (arXiv) are linked below. For the final versions, please consult the journals.

• A PAC-Bayes oracle inequality for sparse neural networks with Mathias Trabs, Springer Proceedings in Mathematics & Statistics, to appear, 2025.
• PAC-Bayes bounds for high-dimensional multi-index models with unknown active dimension, Japanese Journal of Statistics and Data Science, 2025.