The article presents problems connected to the use of neural networks for transforming coordinates from the system "1965" into the system "2000". in order to solve the problem one-directional multi-layer networks have been used, learnt with known optimisation methods: the Levenberg-Marquardt method of conjugate gradients and a quasi -Newton algorithm. In adition, in order to solve the problem of transforming coordinates the Resilient Backpropagation algorithm (RPROP) has been used. The article also presents possibilities of using radial basis networks learnt with the hybrid method with the use of the Gauss function as the base function.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.