The article gives reasoning about the problems that arise in the analysis and use of technologies using neural networks. The greatest attention is paid to the problem of converting known algorithms from the classical description into neural networks. Such problems can be: 1) the need to develop a tool for performing such mathematical operations as addition/subtraction, multiplication/division and others, and also the development of a tool for manipulating data, such as moving, copying, and others, is required; 2) the need to develop structures for management in the calculation (ie, the absence of if, for, while, and others); 3) lack of familiar data structures, such as array, stack, queue and others. Such problems can be demonstrated using the example of finding the smallest path in a graph, it is necessary to relax the edges of a graph, using comparisons and constraints. However, to perform these operations in neural networks, they must first be implemented. It should be noted that the solution of problems of this kind is not typical for neural networks, but it can be part of the realization part of the implementation of the intro-tool for the transformation of classical algorithms into a neural network for the purpose of their joint use, and the explanation of the neural network operation. Also, the proposed approach to solving the problems of implementing a neural system by means of divided functions into several smaller ones, which can be accessed with the help of neural networks. An example of such a partition of the original problem is given.

1. Voevoda A.A., Romannikov D.O. Tezisy o neironnykh setyakh [Theses about neural networks]. Sbornik nauchnykh trudov Novosibirskogo gosudarstvennogo tekhnicheskogo universiteta – Transaction of scientific papers of the Novosibirsk state technical university, 2017, no. 1 (87), pp. 98–108.

2. Bishop C. Pattern recognition and machine learning. Information science and statistics. New York, Springer, 2007. 738 p.

3. Richert W., Coelho L. Building machine learning systems with Python. Birmingham, Packt Publ., 2013. 290 p.

4. Hastie T., Tibshirani R., Friedman J. The elements of statistical learning: data mining, inference, and prediction. 2^nd ed. Springer series in statistics. New York, Springer, 2013. 745 p.

5. Lantz B. Machine learning with R. Birmingham, Packt Publ., 2013. 396 p.

6. Mohri M., Rostamizadeh A., Talwalkar A. Foundations of machine learning. Adaptive computation and machine learning series. Cambridge, MA, MIT Press, 2012. 432 p.

7. Conway D., White J.M. Machine learning for hackers. Sebastopol, CA, O'Reilly, 2012. 324 p.

8. Welcome to the Deep Learning tutorial. Available at: http://deeplear-ning.stanford.edu/tutorial/ (accessed 26.06.2018).

9. Haykin S. Neural networks: a comprehensive foundation. New York, MacMillan Publ., 1994. 1104 p.

10. Voevoda A.A., Romannikov D.O. [Neural networks in systems with sequential actions]. Intellektual'nyi analiz signalov, dannykh i znanii: metody i sredstva [Intellectual analysis of signals, data and knowledge: methods and means], Novosibirsk, 14–17 October 2017, pp. 306–310. (In Russian).

11. Troshina G.V., Voevoda A.A. [Parametric identification of a multi-channel object based on the iterative least-squares method]. 21 Mezhdunarodnaya konferentsiya po myagkim vychisleniyam i izmereniyam (SCM-2018) [21 International Conference on Soft Computing and Measurements (SCM-2018)], St. Petersburg, 23–25 May 2018, pp. 595–598. (In Russian).

12. Troshina G.V., Voevoda A.A. The iterative procedure modeling for the dynamic parameters estimation at the active identification task. Siberian symposium on data science and engineering (2017 SSDSE): proceedings, Akademgorodok, Novosibirsk, 12–13 April 2017, pp. 80–83.

13. Voevoda A.A., Romannikov D.O. Sintez neironnoi seti na osnove seti Petri dlya zadach dvizheniya i stabilizatsii stroya gruppy bespilotnykh letatel'nykh apparatov [Synthesis of a neural network based on the petri network for the tasks of movement and stabilization of the formation of a group of unmanned aerial vehicles]. Vestnik Astrakhanskogo gosudarstvennogo tekhnicheskogo universiteta. Seriya: Upravlenie, vychislitel'naya tekhnika i informatika – Vestnik of Astrakhan State Technical University. Series: Management, Computer Science and Informatics, 2018, no. 2, pp. 26–33.

14. Goodfellow I., Bengio Y., Courville A. Deep learning. Cambridge, MA, MIT Press, 2016.

15. Voevoda A.A., Romannikov D.O. Asinkhronnyi algoritm sortirovki massiva chisel s ispol'zovaniem ingibitornykh setei Petri [Asynchronous sorting algorithm for array of numbers with the use of inhibitory Petri nets]. Trudy SPIIRAN – SPIIRAS Proceedings, 2014, iss. 3 (34), pp. 218?232.

16. Voevoda A.A., Shoba E.V. O razreshimosti zadachi avtonomizatsii mnogokanal'noi sistemy [About diagonally decoupling for multi-input multi-output systems]. Sbornik nauchnykh trudov Novosibirskogo gosudarstvennogo tekhnicheskogo universiteta – Transaction of scientific papers of the Novosibirsk state technical university, 2010, no. 2 (60), pp. 9–16.

17. Voevoda A.A., Markov A.V., Romannikov D.O. Razrabotka programmno-go obespecheniya: proektirovanie s ispol'zovaniem UML diagramm i setei Petri na primere ASU TP vodonapornoi stantsii [Software development: software design using UML diagrams and Petri nets for example automated process control system of pumping station]. Trudy SPIIRAN – SPIIRAS proceedings, 2014, iss. 3 (34), pp. 218–231.

18. Markov A.V., Voevoda A.A. Razvitie sistemy "peremeshchenie manipu-lyatora v prostranstve s prepyatstviyami" pri pomoshchi rekursivnykh funktsii [Development of the system "moving the manipulator in the obstacle space" with the help of recursive functions]. Avtomatika i programmnaya inzheneriya – Automatics & Software Enginery, 2013, no. 2 (4), pp. 35–41.

19. Voevoda A.A., Shoba E.V. Upravlenie perevernutym mayatnikom [Man-agement of the inverted pendulum]. Sbornik nauchnykh trudov Novosibirskogo gosudarstvennogo tekhnicheskogo universiteta – Transaction of scientific papers of the Novosibirsk state technical university, 2012, no. 2 (68), pp. 3–14.