This paper concerns the systems of artificial intelligence because the interest in them is still growing. Neural nets, fuzzy systems and genetic algorithms are successfully used in very different areas like economy, medicine, technology and geology. Those first can be used in all the cases involving prediction, classification and controlling. The fuzzy systems allow problems solving by the differential, simultaneous influence of input variables of a model. However, genetic algorithms have turned out to provide good results in vast spaces. In cause of neural networks it is important, that the minimal condition of their use is certainty (or at least strong suspicion) of the existing dependency between proposed signals (treated as input signal of neural net) and the unknown output signals. The existing information on specific data concerning explored questions is a basis of some impressions concerning essential (or intuitive) choice of input and output signals of a network. The construction of neural net connects both of the signal types making it possible to exploit the model.
Financed by the National Centre for Research and Development under grant No. SP/I/1/77065/10 by the strategic scientific research and experimental development program:
SYNAT - “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.