SYNTHESIS OF NEURONAL ELEMENTS WITH RESPECT TO AN ARBITRARY CHARACTER SYSTEM
Ukrainian Journal of Information Technology · 2026 · 8 (1) · 127-134
DOI: 10.23939/ujit2026.01.127 ↗Kurz gefasst
Fachartikel in der Zeitschrift Ukrainian Journal of Information Technology, 2026. Verfasst von F. E. Geche und 2 weiteren Autor:innen. Die Arbeit nennt 9 Literaturangaben.
Abstract (Originalsprache)
The paper considers a model of a neural element with respect to a given system of characters, which consists of elements of the character group of a multiplicative group that forms the domain of definition of Boolean functions. Such a model of a neural element makes it possible to significantly expand the functional capabilities of neural elements with threshold activation functions, which leads to a simplification of neural networks constructed to implement the required mapping. For the successful use of such neural elements in the construction of neural networks and neuron-like structures, it is necessary to have effective methods for their synthesis. The iterative learning method of a neural element with respect to a given system of characters is defined by a formula that includes the characteristic vector of neurofunctions constructed at the previous iteration step, the structure vector of the current neural element, and a parameter with an admissible domain of values used for training the neural element. Using the developed method, it is always possible to determine whether a given Boolean function is realizable or non-realizable by a single neural element of the specified model. If, during the learning process in the space of characteristic vectors of Boolean neurofunctions, a limit cycle arises and their nontrivial linear combination equals zero, this indicates that the given Boolean function cannot be realized by a single neural element with respect to the given system of characters. Otherwise, the iterative learning process is completed when a structure vector of the neural element that realizes the given function is found. If, with respect to a given system of characters, the function cannot be realized by a single neural element, the system of characters can be modified so that the neural element with respect to the new system of characters will realize the given function. In this paper, based on the iterative method and taking into account the spectral properties of the given Boolean function for determining the system of characters, an efficient algorithm for the synthesis of a neural element that realizes the studied function is developed.
Zitieren
APA
Geche, F. E., Mulesa, O. Y., & Batyuk, A. Y. (2026). SYNTHESIS OF NEURONAL ELEMENTS WITH RESPECT TO AN ARBITRARY CHARACTER SYSTEM. Ukrainian Journal of Information Technology, 8(1), 127-134. https://doi.org/10.23939/ujit2026.01.127
Vancouver
Geche FE, Mulesa OY, Batyuk AY. SYNTHESIS OF NEURONAL ELEMENTS WITH RESPECT TO AN ARBITRARY CHARACTER SYSTEM. Ukrainian Journal of Information Technology. 2026;8(1):127-134. doi:10.23939/ujit2026.01.127
BibTeX
@article{Geche2026,
author = {Geche, F. E. and Mulesa, O. Y. and Batyuk, A. Y.},
title = {SYNTHESIS OF NEURONAL ELEMENTS WITH RESPECT TO AN ARBITRARY CHARACTER SYSTEM},
journal = {Ukrainian Journal of Information Technology},
year = {2026},
volume = {8},
number = {1},
pages = {127--134},
doi = {10.23939/ujit2026.01.127}
}
EUVIMED