Abstract
A mathematical model of the problem of calculating the weight coefficients of a binary neural network is presented. It is proved that in the case of step functions of neuron activation, this model is a system of linear inequalities, which is inconsistent for most practical problems. A method for analyzing the system of inequalities is proposed, which allows calculating the values of the weight coefficients and synthesizing the structure of the neural network, thus ensuring the absolute accuracy of the output signals. The algorithm and an example of implementing the proposed method are given.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
A. V. Kazantsev, “Visual data processing and action control using binary neural network,” in: 8th Intern. Workshop on Image Analysis for Multimedia Interactive Services (WIAMIS’07), Santorini, Greece, 6–8 June 2007, IEEE (2007), p. 23. https://doi.org/10.1109/WIAMIS.2007.90.
S. Liang, R. Sun, Y. Li, and R. Srikant, “Understanding the loss surface of neural networks for binary classification,” in: Proc. 35th Intern. Conf. on Machine Learning (2018), pp. 2835–2843. URL: https://proceedings.mlr.press/v80/liang18a/liang18a.pdf.
M. Krinitskiy, P. Verezemskaya, K. Grashchenkov, N. Tilinina, S. Gulev, and M. Lazzara, “Deep convolutional neural networks capabilities for binary classification of polar mesocyclones in satellite mosaics,” Atmosphere, Vol. 9, No. 11, 426 (2018). https://doi.org/10.3390/atmos9110426.
J. A. Dunnmon, D. Yi, C. P. Langlotz, C. Re, D. L. Rubin, and M. P. Lungren, “Assessment of convolutional neural networks for automated classification of chest radiographs,” Radiology, Vol. 290, No. 2, 537–544 (2019).
S. Korolev, A. Safiullin, M. Belyaev, and Y. Dodonova, “Residual and plain convolutional neural networks for 3D brain mri classification,” arXiv:1701.06643v1 [cs.CV] 23 Jan 2017. URL: https://arxiv.org/pdf/ 1701.06643.pdf.
A. K. Menon and R. C. Williamson, “The cost of fairness in binary classification,” in: Proc. 1st Conf. on Fairness, Accountability and Transparency, PMLR, Vol. 81, 107–118 (2018).
A. Ferreyra-Ramirez, E. Rodriguez-Martinez, C. Aviles-Cruz, and F. Lopez-Saca, “Image retrieval system based on a binary auto-encoder and a convolutional neural network,” IEEE Latin America Trans., Vol. 18, Iss. 11, 1925–1932 (2020). https://doi.org/10.1109/TLA.2020.9398634.
A. Litvinenko, D. Kucherov, and M. Hlybovets, “Decomposition method for calculating the weights of a binary neural network,” Cybern. Syst. Analysis, Vol. 58, No. 6, 889–897 (2022). https://doi.org/10.1007/s10559-023-00522-0.
Author information
Authors and Affiliations
Corresponding author
Additional information
Translated from Kibernetyka ta Systemnyi Analiz, No. 3, May–June, 2024, pp. 15–25; https://doi.org/10.34229/KCA2522-9664.24.3.2.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Litvinenko, A. An Algebraic Method for the Synthesis of Error-Free Binary Neural Network. Cybern Syst Anal 60, 350–358 (2024). https://doi.org/10.1007/s10559-024-00676-5
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10559-024-00676-5