7 Regression Analysis and Parameter Estimation

7.1 INTRODUCTION AND THE GENERAL METHOD OF LEAST SQUARES

The problem of regression analysis is to find coefficients (parameters) of a function that are believed to properly represent a set of experimental data and to perform statistical analyses to confirm (or deny) that the function gives a good fit to the data. Without the additional statistical analysis, just finding the parameters of some candidate function is called curve fitting. Curve fitting, while useful in certain circumstances, is not as powerful as regression analysis. Consider data as represented in Figure 7.1.

It is assumed that x, the ...

Get Numerical Methods for Chemical Engineers Using Excel, VBA, and MATLAB now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.