Institute of Statistics

Workgroup "Dependencies"

• Prof. Dr. Holger Dette
  
• Josua Gösmann
  
• Florian Heinrichs
  
• Kevin Kokot
  

Description:

The modeling of dependence relations is one of the most important subjects in probability theory and statistics. Not surprisingly, a great variety of concepts for dependence structures has emerged, the most popular one being based on second moments of the underlying random variables: the covariance. Nevertheless, it is well known that only linear dependence can be captured by the covariance and that it is characterizing only for a few special classes of distributions, e.g. the multivariate normal distribution.

As a beneficial alternative, the concept of copulas has drawn a lot of attention in the last decades. The copula of a random vector is a distribution function on the unit cube that allows to separate the effect of dependence from the effects of the marginal distributions and therefore completely characterizes stochastic dependence.

The workgroup on dependencies deals with statistical inference within the framework of copulas, including estimation, testing issues and bootstrap approximations.