Abstract
The simplest modular neural network is investigated, in which each module is a single-layer perceptron (a neural kernel). Numerical indices of quality by the criteria of operating speed and plasticity are proposed to characterize structural properties of nuclear neural networks. Based on the indices introduced, optimum procedures of structural synthesis are constructed.
Similar content being viewed by others
REFERENCES
A. Yu. Dorogov andA. A. Alekseyev, “Mathematical models of fast neural networks,” in: Control Systems and Information Processin Izv. GETU, St. Petersburg (1996), pp. 79–84.
A. Yu. Dorogov, “Structural synthesis of fast two-layer neural networks,” Neurocomputers, No. 1 (1999).
H. B. Dwight, Tables of Integrals and Other Mathematical Data [in Russian], Vysshaya Shkola, Moscow (1977).
A. Yu. Dorogov, “Structural models and topological design of fast neural networks,” in: Proc. Intern. Conf. on Information Tools and Technologies, Vol. 1, MIFI, Moscow (1977), pp. 264–269.
A. Yu. Dorogov andA. A. Alekseyev, “Neural networks with nuclear organization,” Oboron. Tekhnika, No. 7–8, 43–46 (1998).
A. Yu. Dorogov andA. A. Alekseyev, “Training of fast neural networks,” in: Information Technologies in Technical and Organizational System, 514 Izv. GETU, St. Petersburg (1997), pp. 68–75.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Dorogov, A.Y. Structural Synthesis of Modular Weakly Connected Neural Networks. II. Nuclear Neural Networks. Cybernetics and Systems Analysis 37, 470–477 (2001). https://doi.org/10.1023/A:1012717318251
Issue Date:
DOI: https://doi.org/10.1023/A:1012717318251