In: Proceedings of the I National Conference on neural networks and applications, Kule, April 1994, pp. 17-28
The ultimate goal of cognitive sciences is to understand how the mind works and the ultimate goal of neural modeling is to build the artificial mind. Short summary of the state of art in this field is given. The symbolic/neural points of view are complementary rather than exclusive. The Floating Gaussian Model (FGM) introduced in this paper facilitates both neural and symbolic interpretations: supervised and unsupervised learning and self-organization of knowledge on the neural side, direct modeling of conceptual space, learning from general laws and retrieval of facts via searching procedures on the symbolic side.
Paper in PDF format, 70 KB
Projects on similar subject and BACK to the on-line publications of W. Duch.