2016:26 Ensemble for deterministic sampling with positive weights

The aim of this project is to find a reliable method for determining an ensemble for Deterministic Sampling with only positive weights. This report is intended to work as a presentation of the project, as well as to work as a beginner’s guide to DS.

In the report several methods for generating DS ensembles are presented, which can encode up to four moments and represent any kind of probability distribution, and encode correlations between parameters. This can be used with many parameters, and the number of samples needed scale linearly with the number of parameters, at best. DS seems to have potential for uncertainty quantification at SSM, in the nuclear power industry and with heavy calculations done in other field.