Abstract:
The article presents an approach to configuration of an artificial neural network architecture and a training set size. Configuration is based on parameter minimization with constraints specifying neural network model quality criteria. The algorithm of artificial neural network architecture and training set size configuration is applied to dynamic object artificial neural network approximation. Series of computational experiments were performed. The method is applicable to construction of dynamic object models based on non-linear autocorrelation neural networks.
Keywords:
dynamic object model, training set, artificial neural network, architecture, training, optimization of artificial neural network architecture.
Received: 03.02.2015
Document Type:
Article
UDC:
004.8, 004.94
Language: Russian
Citation:
A. G. Shumixin, A. S. Boyarshinova, “Algorithm of artificial neural network architecture and training set size configuration within approximation of dynamic object behavior”, Computer Research and Modeling, 7:2 (2015), 243–251
\Bibitem{ShuBoy15}
\by A.~G.~Shumixin, A.~S.~Boyarshinova
\paper Algorithm of artificial neural network architecture and training set size configuration within approximation of dynamic object behavior
\jour Computer Research and Modeling
\yr 2015
\vol 7
\issue 2
\pages 243--251
\mathnet{http://mi.mathnet.ru/crm183}
\crossref{https://doi.org/10.20537/2076-7633-2015-7-2-243-251}
Linking options:
https://www.mathnet.ru/eng/crm183
https://www.mathnet.ru/eng/crm/v7/i2/p243
This publication is cited in the following 2 articles:
V. G. Nikitaev, N. N. Tupitsyn, A. N. Pronichev, V. V. Dmitrieva, E. V. Polyakov, K. A. Liberis, M. S. Grigorieva, A. D. Paladina, “ANALYSIS OF BIOLOGICAL OBJECTS BY DIGITAL OPTICAL MICROSCOPY USING NEURAL NETWORKS”, Bull. Lebedev Phys. Inst., 48:10 (2021), 332
A. G. Shumikhin, A. S. Aleksandrova, “Identifikatsiya upravlyaemogo ob'ekta po chastotnym kharakteristikam, poluchennym eksperimentalno na neirosetevoi dinamicheskoi modeli sistemy upravleniya”, Kompyuternye issledovaniya i modelirovanie, 9:5 (2017), 729–740