ABSTRACT
This chapter discusses different model transformations which can be classified into three groups: model decomposition, model combination, and changes in the output variable. The expression “model decomposition” refers to the extraction of meaningful submodels from a given initial formulation. The chapter presents an example for obtaining the deterministic and stochastic subsystems implied by a transfer function. The expression “model combination” refers to the composition of different models into a single state-space formulation. In practice, model combination is a useful technique for implementing a complex model by defining simpler pieces and then joining them together. The expression “change of variables in the output” refers to a situation where the model output cannot be observed directly, usually because of certain imperfections in the data, such as observation errors, missing values, or aggregation constraints.
