In the theoretical part structural properties of time series (ARCH and GARCH models) and generalized linear models are investigated. The time series mentioned play an important role in the analysis of financial and economical data. Generalized linear models are applicable to polytomous data and are able to handle problems of overdispersion. They can include random effects or can be constructed as nested designs.
Practical Problems are worked out in cooperation with public and private institutions. We are interested e.g., in questions of air pollution, biometry, demand of electrical energy, and robust design of computer chips.