Vine copulas are popular models for dependence modeling. A common modeling assumption in vine copula models is the simplifying assumption, which states that the copulas of conditional distributions are constant with respect to the variables they are conditioned on. This thesis proposes tests to check, whether the simplifying assumption is reasonable for a given data set and to adjust simplified vine copula models if this is not the case. After stating algorithms for simulating from, and evaluating the likelihood of, a non-simplified vine copula, the approach for testing for the simplifying assumption and adjusting simplified vine copula models based on noise contrastive estimation is explained. A simulation study is conducted to check the performance of the suggested tests, before the approach is applied to real data sets.      «

Vine copulas are popular models for dependence modeling. A common modeling assumption in vine copula models is the simplifying assumption, which states that the copulas of conditional distributions are constant with respect to the variables they are conditioned on. This thesis proposes tests to check, whether the simplifying assumption is reasonable for a given data set and to adjust simplified vine copula models if this is not the case. After stating algorithms for simulating from, and evaluat...     »