| Article ID: | iaor20011599 |
| Country: | Cuba |
| Volume: | 21 |
| Issue: | 3 |
| Start Page Number: | 236 |
| End Page Number: | 246 |
| Publication Date: | Sep 2000 |
| Journal: | Revista de Investigacin Operacional |
| Authors: | Zelaya Javier Trejos, Arias Mario Alberto Villalobos |
| Keywords: | neural networks |
We present an algorithm that uses the technique of simulated annealing, for the global optimization problem in non-linear regression. We show the results obtained over some examples, which are comparable to the results obtained with classical methods. We have also implemented the new method in a portable software, that runs on Windows 95 or higher, and is very easy to use. This software contains the simulated annealing for non-linear regression as well as the classical linear regression models.