next up [*] print clean
Next: MULTIVARIATE LEAST SQUARES Up: Table of Contents

Model fitting by least squares

 

The first level of computer use in science and engineering is ``modeling." Beginning from physical principles and design ideas, the computer mimics nature. After this, the worker looks at the result and thinks a while, then alters the modeling program and tries again. The next, deeper level of computer use is that the computer itself examines the results of modeling and reruns the modeling job. This deeper level is variously called ``fitting" or ``inversion." The term ``processing" is also used, but it is broader, including the use of adjoint operators (as discussed in chapter [*]). Usually people are more effective than computers at fitting or inversion, but some kinds of fitting are more effectively done by machines. A very wide range of methods comes under the heading of ``least squares,'' and these methods are the topic of this chapter and chapters [*] through [*].

A part of basic education in mathematics is the fitting of scattered points on a plane to a straight line. That is a simple example of inversion, a topic so grand and broad that some people think of learning to do inversion as simply ``learning.'' Although I will be drawing many examples from my area of expertise, namely, earth soundings analysis, the methods presented here are much more widely applicable.



 
next up [*] print clean
Next: MULTIVARIATE LEAST SQUARES Up: Table of Contents
Stanford Exploration Project
10/21/1998