Discrimination of dynamical system models for biological and chemical processes

In technical chemistry and biotechnology the construction of predictive models has become an essential step in process design and product optimization. Accurate modelling of the reactions in a reactor requires detailed knowledge about the processes involved. However, if, f.e., the development of new products and new production techniques is concerned, this knowledge often is not available. It thus is a typical situation that one has to work with a selection of possible models and the main tasks of early development is to discriminate these models.

Model discrimination means the ranking of models wrt. their ability to reproduce certain experimental data for the process under investigation. In this talk we present a new statistical approach to model discrimination that ranks models wrt. the probability with which they reproduce the data. The talk will shortly review some other prominent approaches in the field, introduces the new approach, discusses its statistical background, presents numerical techniques for its implementation and illustrates the application to realistic examples from biokinetics.