Sensitivity Analysis and Uncertainty Quantification

Abstract

Introduce the idea to use a model for understanding processes or to make predictions. Example where models are used in decision making processes: Political interventions to mitigate the spread of SARS-CoV-2 infections in Germany. Point out, that models are always only an approximation of reality. Therefore, cogency of results should be carefully examined. Part of this examination is an Uncertainty Quantification and Sensitivity Analysis. The structure of this thesis is as follows: Chapter 1 contains a general introduction into mathematical modelling and presents the specific model which is about to be analysed in the following parts. In Chapter 2, different methods for Sensitivity Analysis, Uncertainty Quantification and Parameter Estimation are presented and discussed which of the methods are best to be applied for the model at hand. Chapter 3 then presents a specific model for infection spread prediction and discusses which of the methods of Chapter 2 are most suitable to be applied. After a first analysis of sensitivity and uncertainty, as important identified factors are newly estimated in order to reduce the final output uncertainty. The results are contained in Chapter 4 and concluded in Chapter 5.