3chapter3-1.pptVIP

Chapter 3 Model Fitting Introduction There are three possible tasks when analyzing a collection of data points: 1. Fitting a selected model type or types to the data. 2. Choosing the most appropriate model from competing types that have been fitted. For example, we may need to determine whether the best-fitting exponential model is a better model than the best-fitting polynomial model. 3. Making predictions from the collected data. Relationship BetweenModel Fitting and Interpolation In Task 1 of the preceding paragraph, the precise meaning of best model must be identified and the resulting mathematical problem resolved. In Task 2, a criterion is needed for comparing models of different types. In Task 3, a criterion must be established for determining how to make predictions in between the observed data points. The modeler tends to emphasize the proposed models over the data when model fitting, whereas he or she places greater confidence in the collected data when interpolating and attaches less significance to the form of the model. In a sense, explicative models are theory driven, whereas predictive models are data driven. A modeler may find it necessary to fit a model and also to interpolate in the same problem. The best-fitting model of a given type may prove to be unwieldy, or even impossible, for subsequent analysis involving operations such as integration or differentiation. In such situations the model may be replaced with an interpolating curve (such as a polynomial) that is more readily differentiated or integrated. Sources of Error in the Modeling Process Before discussing criteria on which to base curve-fitting and interpolation decisions, we need to examine the modeling process to ascertain where errors can arise. The errors can be classified under the following category scheme: 1. Formulation error; 2. Truncation error; 3. Round-off error; 4. Measurement error. Formulation errors result from the assumption that certain variables ar