engleski

Localization of the least squares estimate for two parameter regression models

Numerical methods for minimizing the sum of squares require an initial approximation which is as good as possible. Using the concept of the level set we develop a new method for localization of the least squares estimate for two-parameter regression models. Using this method, we suggest a choice of the suitable initial approximation of the parameters. Numerical examples are provided illustrating the effectivness of the author's method. The extension to higher dimensional models is possible.

nonlinear least squares; two parameter regression model; least squares estimate; initial approximation

nije evidentirano