Abstract
The purpose of this paper is to describe a neural network (SNN), that is based on Shannons ideas of reconstruction of a real continuous function from its samples. The basic function, used in this network, is the Sinc-function. Two learning algorithms are described. A simple one called IM-learning and a more complex one based on least square optimization. These two learning methods are illustrated by an example.
Originalsprog | Dansk |
---|---|
Status | Udgivet - 1998 |