Bayesian uncertainty quantication in complex models
The aim of this project is to determine in which cases uncertainty statements resulting from a Bayesian statistical analysis can be trusted.
- 2016 - 2019
Bayesian uncertainty quantification is widely used in practice in many dif- ferent fields of applications, for instance in genetics, finance, epidemiology and machine learning. However, in certain cases it can be fundamentally misleading. The aim of this project is to determine in which cases Bayesian uncertainty statements can be trusted.