ABSTRACT

Mathematical and simulation models are used extensively in many areas of science and industry from population genetics to climate modeling and from simulating a factory production line to theories of cosmology. Modeling may be undertaken for a number of reasons but the most common aim is to predict the behavior of a system under future circumstances. A model may be purely predictive or it may be part of a decision making process by predicting the system behavior under alternative decision scenarios. There are other occasions when a model is just descriptive, simply summarizing the modeler’s understanding of the system (Jeffers 1991). The understanding of the system gained by the modeler and the user can also be an important benefit of the project (Fripp 1985), particularly in scientific research when it can be the principal objective. Equally, a modeling project may have other objectives such as helping to design experiments or identifying research requirements. Despite the great variation in the types of model and their usage, the modeling process itself will take a similar form for most projects and can typically

CONTENTS

2.1 Introduction .................................................................................................. 31 2.2 Choice of the Best Model ............................................................................ 32 2.3 Model Performance .....................................................................................33 2.4 Level of Detail and Complexity ................................................................. 37 2.5 Measuring Model Complexity ...................................................................40 2.6 Relationship between Model Performance and the Level of Detail

or Complexity of a Model ...........................................................................42 2.7 Simplification and Other Related Areas ...................................................46 2.8 Experiment on Model Characteristics and Performance ....................... 47 2.9 Conclusions ................................................................................................... 52 Acknowledgments ................................................................................................53 References ...............................................................................................................53