Theory of continued fraction interpolation and its application in non-linear regression

In fact, when it comes to the problem of the relation between variables is non-linear , we can not use linear regression to solve it. We have to use non-linear regression to analysis as to find the resolution. And it’s a good method to transform the relation non-linear to that of linear relation. In this paper, a method of rational interpolation, continued fraction interpolation is used to solve the non-linear regression problems. And we find it better than the traditional method of Limit Square Method. Finally, we give examples to illustrate its validity.