Is There a Best Model?

Mathematical models are extensively used to supportdecision-making in many disciplines. Neverthelessthere are not any clear standard guidelines to assessmodels performance. This significantly affects modelselection processes, which aim to determine the 'bestmodel', among several possible candidates. Modelperformance is often measured by the accuracy withwhich models predictions fit independentobservations, however this only assesses a singleaspect of a model. The objective of this book is tointroduce a multi-aspect comparison approach whichintends to provide a comprehensive assessment ofmodels capabilities. The innovative aspect of thisapproach is to create a relationship among fourconventional tests, i.e. uncertainty and sensitivityanalysis, goodness-of-fit prediction-observations,model complexity and level of details, in order toprovide a reliable estimation of the differencesbetween constructed and conceptual models. This bookcan be of particular interest to modellers andstudents interested in having an overview of modeltesting as well as scientists willing to applyradioecological models to predict the contaminationof food products.