Towards Reliable Modelling and Learning under Complex Uncertainty, by Thomas Augustin, U. of Munich

ABSTRACT: Imprecise probabilities claim to provide a substantially broader understanding of uncertainty The talk discusses how far this also offers new avenues for reliable statistical modeling and learning After an informal introduction to imprecise probabilities, we consider three prototypic areas the use of neighborhood models as a superstructure upon robust statistics, the expressive modeling of prior data conflict in generalized Bayesian inference and the proper handling of data imprecision.

SPEAKER: THOMAS AUGUSTIN, Department of Statistics at the University of Munich (LMU)
Thomas Augustin is Professor of Statistics at the University of Munich (Germany There he is Head of the “Foundations of Statistics” Group and Dean of Study for the Bachelor and Master Programs in Statistics and in Data Science Thomas has a Diploma and a Ph.D. in Statistics from LMU Munich and has worked at LMU and Bielefeld University before he became Professor at LMU in 2003 His primary research interests are related to the foundations of Statistics and Data Science, including imprecise probabilities, probabilities, (measurement error models, classification under complex uncertainty, decision theory and history of statistical thinking He is also interested in issues of data quality in empirical social research and official statistics.


15:00 – 16:00 PM
Zoom link: