employee
student
student
This article is devoted to current methods of using neural networks to solve optimization problems. The direct economic benefits of their use are highlighted: neural networks can significantly reduce costs, increase resource efficiency, and maximize key financial indicators. Their main advantages stand out: the ability to work with huge amounts of data, find complex, nonlinear relationships, and adapt to a changing environment. Various architectures of neural networks are being investigated, including multilayer perceptrons, precise, recurrent and combined models. Special attention is paid to learning processes and regularization methods. Examples of successful implementation in logistics, finance and industry are given. The effectiveness of neural network approaches and their potential for further development in practical areas is emphasized.
neural networks, gradient descent, machine learning, artificial intelligence, stochastic optimization, multi-criteria tasks, model learning, deep learning, adaptive algorithms
1. Antonov, V. V. Intellektual'nyy metod podderzhki prinyatiya resheniy v tipovoy situacii / V. V. Antonov, K. A. Konev // Ontologiya proektirovaniya. – 2021. – T. 11, No 1(39). – S. 126–136. – DOIhttps://doi.org/10.18287/2223-9537-2021-11-1-126-136.
2. Segaran, T. Programmiruem kollektivnyy razum. – Per. s angl. / T. Segaran. – SPb.: Simvol-Plyus, 2008. – 368 s.
3. Diligenskiy, N. V. Nechetkoe modelirovanie i mnogokriterial'naya optimizaciya proizvodstvennyh sistem v usloviyah neopredelennosti: tehnologiya, ekonomika, ekologiya / N. V. Diligenskiy, L. G. Dymova, P. V. Sevast'yanov. – M. : Izd-vo Mashino-stroenie – 1, 2004. – 397 s.
4. Yarushkina, N. G. Intellektual'nyy analiz vremennyh ryadov. / N. G. Yarushkina, T. V. Afanas'eva, I. G. Perfil'eva. – Ul'yanovsk : UlGTU, 2010. – 320 s.
5. Gladkov, L. A. Geneticheskie algoritmy / L. A. Gladkov, V. V. Kureychik, V. M. Kureychik. – 2010.
