Peter Filzmoser

Peter Filzmoser is a Professor at the Institute for Statistics and Mathematical Economics (E105) of Vienna University of Technology, leading the Computational Statistics Research Area (E105-06) and affiliated with the Network Lab.

His research focuses on:

• Compositional Data Analysis
• Robust Statistics
• Outlier Detection
• Machine Learning for High-Dimensional Data

with applications in geochemistry, mobility data, tribology, and sustainable development.

Recent publications (2023) demonstrate significant advances in explainable outlier detection using Shapley values, robust techniques for compositional data analysis, and applications in forecasting heterogeneous time series. His work extends compositional data analysis through graph signal processing and develops novel robust methodologies for real-world problems.

Professor Filzmoser has advised over 15 Master's and PhD students from 2021-2023. Key research projects he leads include:

• Automotive Intelligence for/at Connected Shared Mobility
• CSTAT: Blind Source Separation
• Generalized relative data and Robustness in Bayes spaces