FORECASTING THE CHINA RMB RATE BASED ON GENETIC ALGORITHM OF NEURAL NETWORKS
Abstract
In recent years, the use of emerging technologies such as artificial intelligence and machine learning has opened up new opportunities in the field of currency exchange rate forecasting. When solving specific practical tasks, the problem of choosing methods that would give the most accurate forecasts arises. Difficulties in obtaining accurate forecasts include unexpected changes in key economic factors. The application of time series forecasting methods in combination with machine learning technologies and neural networks for forecasting exchange rates has become one of the hottest areas of research in recent years, especially with the use of genetic algorithms.Тhis article explores the use of a genetic algorithm in combination with neural networks to forecast the exchange rate of the Chinese yuan. To this end, the authors describe the mechanism of a genetic algorithm, which includes eight steps: initialization, conformity assessment, selection, crossing, mutation, substitution, repetition and selection of the best item. A genetic algorithm is used to optimize the parameters of neural networks, which allows for more accurate forecasts. The authors conducted experiments on historical data of the yuan exchange rate and compared the results with the actual values. The implementation of the genetic algorithm for training the neural network was carried out using the Python programming language. To train the model, two variants of the genetic algorithm were used: selection of the neural network architecture and enumeration of hyperparameters: population size or total number of bots, mutation coefficient, number of survivors and number of iterations. Neural networks of linear and nonlinear architecture were used for training and data forecasting. The results obtained showed that the proposed approach is able to achieve high accuracy in forecasting the Chinese yuan exchange rate. This study may be useful for traders and investors who are interested in forecasting exchange rates to make informed financial decisions.
References
Gulyanitsky L.F., Pavlenko A.I. Development and research of genetic algorithms for forecasting time series. UsiM. 2015. N 3. P.21-29. (in Russian).
Klimenko D.N. Artificial neural network in forecasting currency fluctuations in the stock market. Eurasian Scientific Association. 2021. N 6-4(76). P. 289-292. EDN PYVZDG. (in Russian).
Lomakin N.I., Maksimova O.N., Ekova V.A., Gavrilova O.A., Vagina V.E. Neural networks to predict the value of the dollar using astrological cyclic index Gyushonand Ganu. International Journal of Applied and Fundamental Research. 2016. N 6. P.133-136. (in Russian).
Melikov E.M., Safonova L.A. Forecasting the exchange rate using neural network analysis. Siberian Financial School. 2013. N 5(100). P. 71-76. (in Russian).
Trunev A.P. Forecasting exchange rates using astronomical data using an artificial intelligence system. Scientific journal of KubSAU. 2009. N 51. (in Russian).
Shavshukov V.M., Vorontsovsky A.V., Vyunenko L.F. Analyzing dynamics and forecasting real effective exchangerates for BRICS countries (1994-2016). St Petersburg University Journal of Economic Studies. 2018. N 4(34). P. 568-590. DOI: 10.21638/spbu05.2018.405. EDN PORZHV.
Ermolaev M.B., Golubeva P.A., Mochalova Yu.A. Forecasting exchange rates using adaptive models.Collection of scientific works of Russian universities "Problems of economics, finance and production management". 2023. Is.52. P.164-168. EDN IIJMUS. (in Russian).
Binder A.I., Kononov A.Yu. Yuan in economic and mathematical measurement. Management of economic systems: electronic scientific journal. 2013. N 8(56). P. 1. EDN RKQXUD. (in Russian).
Astrakhantseva I., Kutuzova A., Astrakhantsev R. Artificial Neural Networks in Inflation Forecasting at the Meso-Level.SHS Web of Conferences: III Internationalon New Industrialization and Digitalization (NID 2020). Ekaterinburg: EDP Sciences. 2021. P. 02005. DOI: 10.1051/shsconf/20219302005. EDN HXSSQJ.
Astrakhantseva I.A., Kutuzova A.S., Astrakhantsev R.G. Recurrent neural network for regional inflation forecast. Scientific works of the Free Economic Society of Russia. 2020. Vol. 223. N 3. P. 420-431. DOI: 10.38197/2072-2060-2020-223-3-420-431. EDNQAHOYN. (in Russian).
Astrakhantseva I.A., Gerasimov A.S., Astrakhantsev R.G. Forecasting regional inflation by machine learning algorithms. Ivecofin. 2022. N 4(54). P. 6-13. DOI: 10.6060/ivecofin.2022544.620. EDN ITYDFE. (in Russian).
Genetic algorithm. Simply about the complex. https://habr.com/ru/articles/128704/. (in Russian).
Genetic algorithms. https://algolist.manual.ru/ai/ga/index.php. (in Russian).
Panchenko T.V. Genetic algorithms: educational manual. Astrakhan: "Astrakhan University". 2007. 87 p. (in Russian).
The role of genetic algorithms in modeling. https://habr.com/ru/articles/693742/. (in Russian).
Zaginailo M.V., Fathi V.A. Genetic algorithm as an effective tool for evolutionary algorithms. Innovations. Science. Education. 2020. N 22. P. 513-518. (in Russian).
Nielsen E. Practical Time Series Analysis: Prediction with Statistics and Machine Learning. M.: Dialectics. 2021. 544 p. (in Russian).
Mirolyubova A.A., Ksenofontova O.L. Architecture of neural networks for predicting the development of coronaviral infection. Materials of the III International Scientific- practical conference «Consequences and challenges of the coronavirus pandemic for the technological and socio-economic development of society». Yaroslavl: Yaroslavl State Technical University. 2020. P. 375-380. EDN LWOORZ. (in Russian).
Dynamics of the official exchange rate of a given currency. https://clck.ru/3677fP. (in Russian).
