High-Dimensional Time Series
Concentration results, covariance and precision matrix estimation, factor models, count time series, and change-point analysis for data with temporal and cross-sectional dependence.
Research
Statistical theory for high-dimensional, dependent, and functional data
My research program connects probability, asymptotic theory, and statistical methodology for modern data with complex dependence structures.
Functional and Hilbert-Space Methods
Limit theorems and sample autocovariance theory for functional and Hilbert space-valued processes, with a focus on long-range dependence and nonstationarity.
Extremes and Applications
Extreme value theory and nonlinear dynamics, motivated by applications in econometrics, neuroscience, chemistry, ecology, and other data-rich scientific domains.
Current Direction
Building an independent agenda at TUM
At TUM, I am building a research agenda around statistically principled methods for high-dimensional, temporal, and functional data. A central goal is to develop theory that remains useful for applied scientific questions, especially when observations are dependent, nonstationary, or structurally constrained.
Current themes include discrete data modelling, machine learning, nonlinear dynamics, dimension reduction, change-point analysis, multivariate long-range dependence, nonstationary data, and extreme value theory.
Research Output
Publications and preprints
My publication list includes work on high-dimensional time series, long-range dependence, functional data, change-point analysis, and applications across scientific domains.