Author Guidelines RSCI

 

To participate in the conference, it is necessary to send the texts of articles and the registration data through official conference site on IEEESiberia platform. The deadline is 20 October 2026.

The working languages at the conference are Russian and English.

The originality of the article texts should be at least 70% (according to the website antiplagiat.ru).

The reports are selected by the organising committee for inclusion in the conference agenda and publication in the conference proceedings, the refusal is not explained.

By the beginning of the conference, an electronic version of the conference proceedings will have been issued with the assigned international ISSN number and UDC / LBC codes.

At the end of the conference, the proceedings will be posted in the Scientific Electronic Library (eLIBRARY.RU) and included in the Russian Science Citation Index (RSCI), as well as sent to the conference participants by e-mail, the hard copies (if applicable) will be sent by post at the indicated addresses.

The selected reports will be recommended by the organising committee for publication in the open access journal “Models, Systems, Networks in Economics, Engineering, Nature and Society” (Russian State Commission for Academic Degrees and Titles https://mss_eng.pnzgu.ru).

There is no organizational fee for ordinary participation RSCI publicipation

GUIDELINES FOR SUBMISSION AND PUBLICATION

1. The article should be submitted in A4 format 210 x 297 mm with margins: left margin – 25 mm, right, top and bottom margins – 15 mm. The pages should not contain numbering. The text is drafted in Microsoft Word editor and typed in Times New Roman font, size 14 and 1 interline spacing.

2. At the beginning of the article, the UDC index is indicated. On the next line the title in Russian is printed in capital letters, the font is bold. Below, with 1,5 interline spacing the surname, name, patronymic of the author(s) age given in lowercase letters. Then with 1,5 interline spacing the full name of the organisation / institution, city and country is given. Further, the same data are provided in English. After that, the abstract and keywords in Russian and English are placed.

3. The text of the article may contain formulas, tables and figures. The references are drawn up in accordance with the Russian national references standard GOST R 7.0.100-2018.

4. The text of the article in Russian or English with a volume of 3-4 pages should be given in a separate file named by the name of the first author and the first three words of the title of the article.

Should you have any questions, please contact the secretary of the organising committee

Drozhdin Vladimir Viktorovich (tel: +7 937 429 79 40, e-mail: probinf@yandex.ru)
website: http://probinf.ru/

 

SAMPLE ARTICLE

УДК 004.9

МОДЕРНИЗИРОВАННЫЙ АЛГОРИТМ ОБУЧЕНИЯ ВЕСОВ РАДИАЛЬНЫХ БАЗИСНЫХ НЕЙРОННЫХ СЕТЕЙ
ПРИ РЕШЕНИИ КРАЕВЫХ ЗАДАЧ

             В.И. Горбаченко, Е.В. Артюхина

Пензенский государственный университет,
г. Пенза, Россия

Modernized algorithm of training the scales of radial basis neural networks by solving boundary equations

                Gorbachenko V.I., Artyukhina E.V.

Penza State University, Penza, Russia

 

Аннотация: Разработан модернизированный алгоритм обучения весов радиальных базисных нейронных сетей при решении краевых задач. Экспериментально показано, что данный алгоритм позволяет сократить время решения задачи по сравнению с алгоритмом сопряженных градиентов для минимизации квадратичного функционала.

Ключевые слова: радиальная базисная сеть, нейронная сеть, алгоритм обучения

 

Abstract. Modernized algorithm of training the scales of radial basis neural networks by solving boundary equations is worked out. It is experimentally proved that this algorithm allows of reduction the time for solving the tasks in comparison with the algorithm of conjugated gradients for minimization of the quadratic functional.

Keywords: radial basic network, neural network, learning algorithm

 

Radial basic functions neural networks (RBFNNs) find effective application in solving boundary value problems in mathematical physics [1]. The work aims at developing and studying a modernized RBFNN weight training algorithm. We consider learning as illustrated by the Poisson equation

, ,                        (1)

where  is an area boundary; and  are known functions .

 

Figure 1 ‑ Comparison of the algorithms

The results of the experiments show that the developed algorithm reduces the problem solving time by about 20% compared to the conjugate gradient algorithm for minimizing the quadratic learning functional of RBFNN weights.

 

References

 

1. Numerical solution of elliptic partial differential equation using radial basis function neural networks / L. Jianyu, L. Siwei, Q. Yingjiana, H. Yapinga // Neural Networks, 2003, 16(5/6). – P. 729 – 734.

2. Dennis J. Jr., Shnabel R. Numerical methods of unconditional optimization and solution of nonlinear equations / translated from English. – M.: Mir, 1988. – 440 p.

3. Vorst van der, H. Iterative Krylov Methods for Large Linear Systems. – Cambridge: Cambridge University Press, 2003. – 232 p.

4. Artyukhin V.V., Artyukhina E.V., Gorbachenko V.I. Radial-basic neural networks for solving boundary value problems by meshless methods // Scientific session of National Research Nuclear University MEPHI-2010. 12th All-Russian scientific and technical conference “Neuroinformatics-2010”: proceedings. – M.: NRNU MEPHI, 2010, Part 2. – P. 237 – 247.

5. Arbuzova A.A. Diagnosis of pneumonia by X-ray images using convolutional neural networks // Models, Systems, Networks in Economics, Engineering, Nature and Society, 2021, no. 2. – P. 107 – 114