Modeling an experimental system often results in a number of alternative models that are all justified by the available experimental data. In order to discriminate between these models, additional experiments are needed. We present a method for experiment selection that helps in discriminating between differential equation models of experimental systems in a systematic and efficient way. The method generalizes upon previous work on model discrimination in that it deals with semi-quantitative differential equations, which use interval bounds on parameter values and envelopes for functional relations. The model discrimination method is based on an entropy criterion for the selection of the most informative experiment. The applicability of the method to real-world problems is illustrated by means of an example in population biology, the discrimination of competing models of the growth of phytoplankton in a bioreactor.

Full text