V.Y. Ilichev1, I.V. Chukhraev2

1, 2 Kaluga Branch of the Bauman MSTU (Kaluga, Russia)

The article is devoted to the consideration of one of the areas of application of modern and promising computer technology – machine learning. This direction is based on the creation of models consisting of neural networks and their deep learning.

At present, there is a need to generate new, not yet existing, images of objects of different types. Most often, text files or images act as such objects. To achieve a high quality of results, a generation method based on the adversarial work of two neural networks (generator and discriminator) was once worked out. This class of neural network models is distinguished by the complexity of topography, since it is necessary to correctly organize the structure of neural layers in order to achieve maximum accuracy and minimal error.

The described program is created using the Python language and special libraries that extend the set of commands for performing additional functions: working with neural networks Keras (main library), integrating with the operating system Os, outputting graphs Matplotlib, working with data arrays Numpy and others. A description is given of the type and features of each neural layer, as well as the use of library connection functions, input of initial data, compilation and training of the obtained model.

Next, the implementation of the procedure for outputting the results of evaluating the errors of the generator and discriminator and the accuracy achieved by the model depending on the number of cycles (eras) of its training is considered. Based on the results of the work, conclusions were drawn and recommendations were made for the use and development of the considered methodology for creating and training generative and adversarial neural networks.

Studies have demonstrated the procedure for operating with comparatively simple and accessible, but effective means of a universal

Python language with the Keras library to create and teach a complex neural network model. In fact, it has been proved that the use of this method allows to achieve high-quality results of machine learning, previously achievable only when using special software systems for working with neural networks.

Ilichev V.Y., Chukhraev I.V. Data processing using deep learning of the generative-adversarial neural network (GAN). Neurocomputers. 2021. V. 23. № 5. P. 51−56. DOI: https://doi.org/10.18127/j19998554-202105-04 (in Russian)

1. Chumakov V.E. Ispolzovanie tekhnologii DeepLearning. Sb. nauchnykh trudov po materialam Mezhdunar. nauchno-praktich. konf. «Traektoriya nauchno-tekhnologicheskogo razvitiya Rossii s uchetom globalnykh trendov». Pod obshch. red. E.P. Tkachevoi. 2019. S. 172−175. (in Russian)
2. Magomadov V.S. Primenenie generativno-sostyazatelnoi seti dlya sozdaniya sinteticheskikh dannykh. Vestnik sovremennykh issledovanii. 2018. № 6.3(21). S. 517−519. (in Russian)
3. Fedorenko S.A., Feshina E.V. Sfery primeneniya generativno-sostyazatelnykh neirosetei. Sb. materialov I Vseros. studencheskoi nauchno-praktich. konf. «Tsifrovizatsiya ekonomiki: napravleniya, metody, instrumenty». 2019. S. 226−228. (in Russian)
4. Shkirya A.S. Razrabotka generativno-sostyazatelnoi seti dlya sozdaniya izobrazhenii. Sb. statei po materialam XCII studencheskoi Mezhdunar. nauchno-praktich. konf. «Nauchnoe soobshchestvo studentov. Mezhdistsiplinarnye issledovaniya». 2020. S. 37−42. (in Russian)
5. Kingma, Diederik and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980. 2014. URL = https://arxiv.org/abs/1412.6980v8 (data obrashcheniya: 09.07.2021).
6. Demin I.S., Belov Yu.S., Chukhraev I.V. Obuchenie svertochnoi neironnoi seti na baze arkhitektury U-net s ispolzovaniem minimalnykh resursov. Elektromagnitnye volny i elektronnye sistemy. 2019. T. 24. № 7. S. 24−29. (in Russian)
7. Plotnikov M.V. GAN – generativnye sostyazatelnye iskusstvennye neironnye seti. Sb. materialov XIII Vseros. molodezhnoi nauchnoinnovatsionnoi shkoly «Matematika i matematicheskoe modelirovanie». 2019. S. 290−292. (in Russian)
8. Tkachyov N.M., Fedyaev O.I. Parametricheskoe opisanie modelei glubokikh neironnykh setei v biblioteke Keras. Sb. nauchnykh trudov II Mezhdunar. nauchno-praktich. konf. «Programmnaya inzheneriya: metody i tekhnologii razrabotki informatsionno-vychislitelnykh sistem» (PIIVS-2018) (studencheskaya sektsiya). V 2-kh tomakh. Donetsk. 2018. S. 259−263. (in Russian)
9. Keras: The Python Deep Learning library. Keras documentation. URL = https://keras.io/#why-thisname-keras. (data obrashcheniya 09.07.2021).
10. Ilichev V.Yu. Ispolzovanie parsinga dlya sozdaniya bazy meteorologicheskikh dannykh i razrabotka na ee osnove neirosetevoi modeli prognozirovaniya skorosti vetra. Sistemnyi administrator. 2020. № 10(215). S. 92−95. (in Russian)
11. Frolov P.V., Chukhraev I.V., Grishanov K.M. Primenenie iskusstvennykh neironnykh setei v sistemakh obnaruzheniya vtorzhenii. Sistemnyi administrator. 2018. № 9(190). S. 80−83. (in Russian)
12. Dli M.I., Puchkov A.Yu., Lobaneva E.I. Algoritmy formirovaniya izobrazhenii sostoyanii obieektov dlya ikh analiza glubokimi neironnymi setyami. Prikladnaya informatika. 2019. T. 14. № 2(80). S. 43−55. (in Russian)
13. Ilichev V.Yu., Yurik E.A. Obrabotka statisticheskikh dannykh metodom glubokogo obucheniya s ispolzovaniem modulya Keras. Nauchnoe obozrenie. Tekhnicheskie nauki. 2020. № 5. S. 16−20. (in Russian)