Application of a Long Short-Term Memory Network to Predict Stock Prices under Uncertainty
DOI:
https://doi.org/10.52575/2687-0932-2023-50-3-669-680Keywords:
machine learning, deep learning, artificial neural network, financial time series forecasting, stock tradingAbstract
Forecasting the prices of shares of various companies is one of the most important and complex tasks in the financial economy. The higher the forecasting accuracy, the more profit an investor can get from stock trading. Currently, there are many forecasting models, both statistical and using machine learning. However, all these models have one big drawback - they are not able to take into account the data sequence when predicting, which significantly reduces the prediction accuracy. To solve this problem, it was proposed to use a deep learning model a recurrent neural network of long short-term memory. This model is able to take into account the chronology of data, as well as work with a large amount of historical data. The purpose of this study is to analyze and develop a long short-term memory neural network for predicting stock prices. To analyze the proposed forecasting model, experiments were carried out to assess the accuracy of forecasts issued by the developed neural network. As a result of the experiments, it was found that the neural network of long short-term memory, according to the results of forecasting, is superior to the rather popular statistical model SARIMAX. The developed model can be useful for investors as an additional effective tool for forecasting stock prices.
Downloads
References
Губарева Е.А. 2020. Нейронные сети в анализе временных рядов. Инновации и инвестиции, 10: 150-153.
Губарева Е.А., Хашин С.И. 2021. Вопросы эффективности обучения нейронных сетей при анализе временных рядов. Экономика: вчера, сегодня, завтра. 9(1): 121-127.
Имамзазин Т. Р. 2019. Основные методы и модели прогнозирования будущего курса акций. Наука, Техника и Образование, 10(63): 63-67.
Нестерова К.И. 2020. Обзор современных методов прогнозирования динамики цен на фондовом рынке. Наукосфера, 7: 91-95.
Обрубов М.О. 2021. Применение LSTM-сети в решении задачи прогнозирования многомерных временных рядов. Национальная Ассоциация Ученых, 68(2): 43-48.
Передриенко А.И., Лютая Т.П., Харитонов И.М., Степанченко И.В. 2020. Методы краткосрочного прогнозирования финансовых временных рядов с малыми объёмами выборки. ИВД, 5: 65.
Сунчалин А.М. 2020. Обзор методов и моделей прогнозирования финансовых временных рядов. Хроноэкономика, 1: 25-29.
Тимофеев А.Г., Лебединская О.Г. 2022. Модель применения сверточной нейронной сети (cnn) в сочетании с долговременной памятью (lstm) прогнозирования цены на нефть в условиях неопределенности. ТДР, 2: 35-44.
Хоботов В.А. 2021. Анализ тональности финансовых новостей с применением нейросетевых моделей для прогноза динамики цен на нефть марки Brent. Актуальные вопросы современной экономики, 2: 138-143.
Balcl M. A. 2020. Fractional Interaction of Financial Agents in a Stock Market Network. Applied Mathematics and Nonlinear Sciences, 5(1): 317-336. DOI: 10.2478/amns.2020.1.00030
Cui X., Hu J., Wu P. 2021. Investigation of stock price network based on time series analysis and complex network. International Journal of Modern Physics B. DOI: 10.1142/S021797922150171X
Guo W., Li Z., Gao C., Yang Y. 2022. Stock price forecasting based on improved time convolution network. Computational Intelligence. DOI: 10.1111/coin.12519
Khoojine A. S., Han D. 2020. Stock price network autoregressive model with application to stock market turbulence. The European Physical Journal B - Condensed Matter and Complex Systems, 93(7): 133. DOI: 10.1140/epjb/e2020-100419-9
Liu G., Ma.W. 2022. A quantum artificial neural network for stock closing price prediction. Information Sciences, 598: 75-85. DOI: 10.1016/j.ins.2022.03.064
Mezghani T., Ben Hamadou F., Boujelbène Abbes M. 2021. The dynamic network connectedness and hedging strategies across stock markets and commodities: COVID-19 pandemic effect. Asia-Pacific Journal of Business Administration. DOI: 10.1108/APJBA-01-2021-0036
Nayak S.C., Misra B.B. 2018. Estimating stock closing indices using a GA-weighted condensed polynomial neural network. Financial Innovation, 4(1): 21. DOI: 10.1186/s40854-018-0104-2
Nor S.M., Zawawi N.H.M. 2020. A neural network approach for fundamental investment analysis: a case of Athens Stock Exchange. Economic Annals-ХХI, 182(4): 56-63. DOI: 10.21003/ea.V182-07
Tao Z., Chunhui L., Muzhou H. 2018. Forecasting stock index with multi-objective optimization model based on optimized neural network architecture avoiding overfitting. Computer Science and Information Systems, 15(1): 211-236. DOI: 10.2298/CSIS170125042T
Yamin S., Gulzar S. 2020. Multiples and stock price, new approach for relative valuation through neural network. Singapore Economic Review. DOI: 10.1142/S0217590820480045
Yan Z., Zhou K., Zhu. X, Chen H. 2022. Application of MEA-LSTM Neural Network in Stock Balance Prediction. Lecture Notes on Data Engineering and Communications Technologies, 121: 60-71. DOI: DOI 10.1007/978-3-030-97057-4_6
Abstract views: 168
Share
Published
How to Cite
Issue
Section
Copyright (c) 2023 Economics. Information Technologies
This work is licensed under a Creative Commons Attribution 4.0 International License.
