ORTHOGONALIZATION OF NEURAL NETWORK WEIGHT COEFFICIENTS

Zuev V.M

doi:10.24412/2413-7383-135-140

Zuev V.M.

About the authors

Zuev V.M.

Зуев Владимир Михайлович – заведующий отделом ИРТС Федерального государственного бюджетного научного учреждения «Институт проблем искусственного интеллекта». Область научных интересов: интеллектуальные робототехнические системы, адрес: 283048, г. Донецк, ул. Артема, д. 118 б.

UDC: 004.93
DOI: 10.24412/2413-7383-135-140
Language: Russian

Abstract: The article considers the regularization of a convolutional neural network using the orthogonalization method. The proposed method can be used to reduce the number of optimized weight coefficients. A preliminary filter adjustment stage is introduced, which allows choosing the optimal number of convolution layers and speeding up training due to the optimal filter adjustment at the beginning of training. The analysis shows that the method we propose is quite efficient.

Keywords: neural network, regularization, orthogonalization. Articulus disserit regu larizationem retis neuralis convolutionis utendi methodo orthogonalizationis. Proposita methodus adhiberi potest ad reducere numerum coefficientium optimized ponderis. Filtrum

References

Verzhbitskiy V.M. Numerical methods M. Higher School of Economics, 2001
Saad Yu. Iterative methods for sparse linear systems: Study guide. – In 2 volumes. Volume 1 / Translated from English by H.D. Ikramov; Preface by V.A. Sadovnichy, Moscow: Moscow University Press, 2013– 344 p.
https://habr.com/ru/companies/smartengines/articles/495816/4.8K
Shilov G. E. Finite-dimensional linear spaces. Moscow: Nauka Publ. 1969
Beklemishev D.V. Additional chapters of linear algebra. Moscow: Nauka, 1983.

Issue: 1(36)'2025
Section: ROBOTS, MECHATRONICS AND ROBOTIC SYSTEMS
How to cite: Zuev V.M.. ORTHOGONALIZATION OF NEURAL NETWORK WEIGHT COEFFICIENTS // Problems of Artificial Intelligence. - 2025. - № 1 (36). - P. 135-140. - https://paijournal.guiaidn.ru/en/2025/1(36)-11.html


	РУС ENG
	ORTHOGONALIZATION OF NEURAL NETWORK WEIGHT COEFFICIENTS
About the magazine News Goals and sphere Founder and publisher Editorial Board Licensing conditions Confidentiality Attitude towards plagiarism Publication ethics Archiving Policy Subscription For authors Instructions for authors The review process Copyright Agreement on the transfer of rights Editorial fees Archive All issues Search Contacts Contacts	Zuev V.M. About the authors Zuev V.M. Зуев Владимир Михайлович – заведующий отделом ИРТС Федерального государственного бюджетного научного учреждения «Институт проблем искусственного интеллекта». Область научных интересов: интеллектуальные робототехнические системы, адрес: 283048, г. Донецк, ул. Артема, д. 118 б. UDC: 004.93 DOI: 10.24412/2413-7383-135-140 Language: Russian Abstract: The article considers the regularization of a convolutional neural network using the orthogonalization method. The proposed method can be used to reduce the number of optimized weight coefficients. A preliminary filter adjustment stage is introduced, which allows choosing the optimal number of convolution layers and speeding up training due to the optimal filter adjustment at the beginning of training. The analysis shows that the method we propose is quite efficient. Keywords: neural network, regularization, orthogonalization. Articulus disserit regu larizationem retis neuralis convolutionis utendi methodo orthogonalizationis. Proposita methodus adhiberi potest ad reducere numerum coefficientium optimized ponderis. Filtrum References Verzhbitskiy V.M. Numerical methods M. Higher School of Economics, 2001 Saad Yu. Iterative methods for sparse linear systems: Study guide. – In 2 volumes. Volume 1 / Translated from English by H.D. Ikramov; Preface by V.A. Sadovnichy, Moscow: Moscow University Press, 2013– 344 p. https://habr.com/ru/companies/smartengines/articles/495816/4.8K Shilov G. E. Finite-dimensional linear spaces. Moscow: Nauka Publ. 1969 Beklemishev D.V. Additional chapters of linear algebra. Moscow: Nauka, 1983. Issue: 1(36)'2025 Section: ROBOTS, MECHATRONICS AND ROBOTIC SYSTEMS How to cite: Zuev V.M.. ORTHOGONALIZATION OF NEURAL NETWORK WEIGHT COEFFICIENTS // Problems of Artificial Intelligence. - 2025. - № 1 (36). - P. 135-140. - https://paijournal.guiaidn.ru/en/2025/1(36)-11.html
Founder and publisher: State Institution «Institute of Problems of Artificial Intelligence» Address: 83048, Donetsk, Artema st.,118 Editor-in-chief: V. Ju. Shelepov. © State Institution «Institute of Problems of Artificial Intelligence»,