EXPERIMENTAL DETERMINATION OF THE OPTIMAL PARAMETERS OF THE RECURRENT NEURAL NETWORK FOR THE TASKS OF PATENT CLASSIFICATION

Indicators of patent activity are now often used in technological forecasting and competitive intelligence. An important role is to predict the development of patent trends in individual countries and around the world, which allows to identify the main priority directions of technology development. Analysis of patents in the field of analog technology was fulfilled. The international patent classification is outdated, most studies are interdisciplinary. There is a need to select and create new classes. The purpose of this study is to analyze the parameters that affect the results of the recurrent neural network, designed for the thematic classification of the patent array. The analysis of the identified parameters affects the quality of the neural network and the selection of optimal values. The optimal parameters of the neural network were determined: the number of layers, the size of the layer, the value of the exclusion parameter, the batch size for training in the network, the choice of the Keras library optimizer was made. The reported study was funded by RFBR according to the research project № 19-07-01200.

1. Methods of Statistical and Semantic Patent Analysis / D.M. Korobkin, S.A. Fomenkov, A.G. Kravets, S.G. Kolesnikov // Creativity in Intelligent Technologies and Data Science. Second Conference, CIT & DS 2017 (Volgograd, Russia, September 12-14, 2017): Proceedings / ed. by A. Kravets, M. Shcherbakov, M. Kultsova, Peter Groumpos; Volgograd State Technical University [et al.]. - [Germany]: Springer International Publishing AG, 2017. - p. 48-61. - (Ser. Communications in Computer and Information Science; Vol. 754).

2. Determination of the patent and the organization of the rights of the patent holder [Electronic resource]. - Access mode: http://economicdefinition.com/Economic_and_legal_terminology/Patent_Patent__eto.html (circulation date 01.06.2018).

3. Kondratieva TN, Forecasting the tendency of financial time series using the LSTM neural network // Internet journal Naukovedenie. 2017. V. 9. № 4. P. 61.

4. Kravets, A.G. Patents Images Retrieval and Convolutional Neural Network Training Dataset Quality Improvement [Electronic resource] / А.G. Kravets, N.S. Lebedev, M.S. Legenchenko // Proceedings of the IV International Research Conference on Information Technologies in Science, Management, Social Sphere and Medicine (ITSMSSM 2017) / ed. by O.G. Berestneva [et al.]. - [Published by Atlantis Press], 2017. - p. 287-293. - (Ser. Advances in Computer Science Research (ACSR); Vol. 72). - URL: https://www.atlantispress.com/proceedings/itsmssm-17.

5. Kline, D.M., Revisiting squared-error and cross-entropy functions for training neural network classifiers. [Electronic resource]. - 2005. - Access mode: https://link.springer.com/article/10.1007/s00521-005-0467-y (Contact date: 05/04/2019).

6. Lance G. N., Willams W. T. A general theory of classi fi cation sorting strategies. 1. hierarchical systems // Comp. J. - 1967. - no. 9. - Pp. 373–380.

7. Jain A., Murty M., Flynn P. Data clustering: A review // ACM Computing Surveys. — 1999. — Vol. 31, no. 3. — Pp. 264–323.

8. Eck, D., Schmidhuber, J. A First Look at Music Composition using LSTM Recurrent Neural Networks. - Manno, Switzerland: Instituto Dalle Molle di studi sull ’intelligenza artificiale. [Electronic resource]. - 2002. - Access mode: http://people.idsia.ch/~juergen/blues/IDSIA-07-02.pdf (Revised: 05/04/2019).

9. The Python Deep Learning library [Electronic resource]. - Access mode: https://keras.io/ (request date 10.05.2019)

10. "Smart Queue" Approach for New Technical Solutions Discovery in Patent Applications / A.G. Kravets, N. Shumeiko, B. Lempert, N. Salnikova, N.L. Shcherbakova // Creativity in Intelligent Technologies and Data Science. Second Conference, CIT & DS 2017 (Volgograd, Russia, September 12-14, 2017): Proceedings / ed. by A. Kravets, M. Shcherbakov, M. Kultsova, Peter Groumpos; Volgograd State Technical University [et al.]. - [Germany]: Springer International Publishing AG, 2017. - p. 37-47. - (Ser. Communications in Computer and Information Science; Vol. 754).