Performance assessment of the anticipatory approach to optimal experimental design for model discrimination

The problem of model discrimination arises when several models are proposed to describe one and the same process, a situation encountered in many research fields. To identify the best model from the set of rival models, it may be necessary to collect new information about the process, and thus additional experiments have to be performed. Several approaches have been described in literature to design optimal discriminatory experiments. The anticipatory approach is one of them and is very appealing from a conceptual point of view because the expected information content of the newly designed experiment is considered, even before the experiment is performed (anticipatory design). In this paper, the performance of this approach is evaluated by comparing it with the performance of other, established approaches to optimal experimental design for model discrimination. To conduct this comparison four performance measures were defined: (1) whether the most appropriate model could be identified, (2) the number of additional experiments that have to be designed and performed to achieve model discrimination, (3) the quality of the parameter estimates of the model that is eventually identified as the most appropriate one, and (4) the rate at which the inadequate models are identified. The results clearly indicate that the anticipatory approach has its benefits and may be the preferred approach in many applications in (bio)chemical engineering and in-silico biology.