پیش‌بینی رفتار بورس اوراق بهادار با به‌کارگیری اندیکاتورهای تکنیکال، مبتنی بر رویکردهای یادگیری تقویتی عمیق و شبکه‌های کانولوشن مطالعه موردی: بورس اوراق بهادار ایران

هادی‌زاده, آنیتا; تارخ, محمد جعفر; میرزایی قزاآنی, مجید

پیش‌بینی رفتار بورس اوراق بهادار با به‌کارگیری اندیکاتورهای تکنیکال، مبتنی بر رویکردهای یادگیری تقویتی عمیق و شبکه‌های کانولوشن مطالعه موردی: بورس اوراق بهادار ایران

نوع مقاله : مقاله پژوهشی

نویسندگان

¹ دانشجوی دکتری مهندسی فناوری اطلاعات، دانشکده مهندسی صنایع، دانشگاه صنعتی خواجه نصیر، تهران، ایران

² استاد، گروه مهندسی فناوری اطلاعات، دانشکده مهندسی صنایع، دانشگاه صنعتی خواجه نصیر، تهران، ایران

³ استادیار، گروه مهندسی مالی، دانشکده مهندسی صنایع، دانشگاه صنعتی خواجه نصیر، تهران، ایران

چکیده

این مقاله با استفاده از یادگیری تقویتی عمیق مدلی ارائه می‌کند تا وظایف یک معامله‌گر در بازار بورس ایران را با توجه به سهم‌های نقد شونده مدل‌سازی کند. قیمت سهام به همراه اندیکاتورهای مبتنی بر آن به‌عنوان ورودی به شبکه عصبی کانولوشن وارد می‌شوند.سپس، با استفاده از اندیکاتورهای محاسبه شده، داده‌های قیمت بر اساس تاریخ تطبیق داده می‌شود. به منظور محاسبه میزان تطبیق خروجی محاسبه شده با خروجی مورد انتظار، از تابع هزینه‌ی مجموع مربعات خظا استفاده می‌شود که در فرایند بهینه‌سازی کمینه می‌شود. همچنین با به‌کارگیری مدل‌های کانولوشن به جای جداول Q از بیش برارزش مدل به دلیل وجود داده‌های کم برای آموزش مدل جلوگیری به عمل آمده است. از طرفی با استفاده از اطلاعات موجود در حجم معاملات، از این سیگنال به عنوان نقشی مکمل در پیش‌بینی روند آینده سهم‌ها بهره گرفته شده است. و برای ارزیابی، برتری این مدل نسبت به استراتژی خرید و نگهداری مقایسه شده است.

کلیدواژه‌ها

20.1001.1.24766291.1401.7.4.3.0

عنوان مقاله [English]

Forecasting Stock market behavior through implementing technical indicators , based on deep reinforcement learning and convolutional network approaches Case study: Iran Stock Market

نویسندگان [English]

Anita Hadizadeh ¹
Mohammad jafar Tarokh ²
Majid Mirzaee Ghazani ³

¹ PhD student in Information Technology Engineering, Faculty of Industrial Engineering, Khajeh Nasir University of Technology, Tehran, Iran

² Professor, Department of Information Technology Engineering, Faculty of Industrial Engineering, Khajeh Nasir University of Technology, Tehran, Iran

³ Assistant Professor, Department of Financial Engineering, Faculty of Industrial Engineering, Khajeh Nasir University of Technology, Tehran, Iran

چکیده [English]

Nowadays, the stock market plays an important role in the economy of different countries. The abundance of data in the stock market and the need for fast and correct data processing and making appropriate decisions have made the use of computers inevitable. In this article, using deep reinforcement learning, a model is designed to present the duties of a trader in the Iranian stock market with regard to liquidated shares. In the first step, the history of stock prices along with the indicators based on it are given as inputs to the convolutional neural network. In the next step, in order to calculate the matching rate of the convolution output with the expected output, the sum of squared error cost function is used, which, in turn, is minimized in the optimization process. Since the data in the Iranian stock market is limited, using the convolution model instead of the Q table in the deep reinforcement model prevents the over fitting of the model. In order to evaluate the model, the data of Tehran Stock Exchange was used in the period of 1390 to 1400. The performance of the proposed method was compared with the buy and hold strategy. The results show, in some cases, the profit from the proposed method in contrast with buy and hold strategy is 21% and -7%, respectively.

کلیدواژه‌ها [English]

deep reinforcement learning
convolutional network
technical indicator
technical analysis

مراجع

[1] Sharma, A., D. Bhuriya, and U. Singh. Survey of stock market prediction using machine learning approach. in 2017 International conference of electronics, communication and aerospace technology (ICECA). 2017. IEEE.

[2] Sutton, R.S. and A.G. Barto, Reinforcement learning: An introduction. Robotica, 1999. 17(2): p. 229-235.

[3] Barroso, B., R. Cardoso, and M. Melo, Performance analysis of the integration between Portfolio Optimization and Technical Analysis strategies in the Brazilian stock market. Expert Systems with Applications, 2021. 186: p. 115687.

[4] DEHGHAN, A. and M. Kamyabi, On Tehran stock market in cyclical condition. 2019.

[5] Hejazi, R., et al., The Relationship between Stock Prices Crash and Dedicated Institutional Investors and Transient Institutional Investors. 2020.

[6] Ali Asghar, A.R. and B. Lari Semnani, Assessing the Relationships of Bank Deposits and Governmental Industrial Development Bonds Investments with the Attracitveness of Investing (Lequidity and Capitalization) in Tehran Stock Exchange (TSE). Management Research in Iran, 2007. 11(20): p. 1-29.

[7] Fairchild, R.J. and J. Kinsella, An Emotional Finance Framework for Examining Bubbles and Crashes. Available at SSRN 3999323, 2022.

[8] Hatefi Madjumerd, M., G. Zamanian, and M.N. Shahiki Tash, Evaluation of Multiple Bubbles in the Stock Market of Tehran. Quarterly Journal of Quantitative Economics, 2017. 14(2): p. 85-110.

[9] Bagherzadeh, S., The Initial Public Offerings Underpricing and Its Determinants in Tehran Stock Exchange. Human Sciences Modares, 2011. 15(1): p. 77-107.

[10] Afsar, A. and F. Helyel, A Hybrid Approach to Portfolio Optimization Using Technical Analysis and Data Mining. Modern Research in Decision Making, 2017. 2(2): p. 1-22.

[11] Katani, S., F. Samadi, and Z. Hajiha, Optimization of multi-objective portfolio using imperialist competitive algorithm in Tehran Stock Exchange. 2021.

[12] BELAGHİ, R.A., M. AMİNNEJAD, and Ö.G. ALMA, Stock Market Prediction Using Nonparametric Fuzzy and Parametric GARCH Methods. Turkish Journal of Forecasting, 2018. 2(1): p. 1-8.

[13] Nabipour, M., et al., Deep learning for stock market prediction. Entropy, 2020. 22(8): p. 840.

[14] Nabipour, M., et al., Predicting Stock Market Trends Using Machine Learning and Deep Learning Algorithms Via Continuous and Binary Data; a Comparative Analysis. IEEE Access, 2020. 8: p. 150199-150212.

[15] Deng, Y., et al., Deep direct reinforcement learning for financial signal representation and trading. IEEE transactions on neural networks and learning systems, 2016. 28(3): p. 653-664.

[16] Xiong, Z., et al., Practical deep reinforcement learning approach for stock trading. arXiv preprint arXiv:1811.07522, 2018.

[17] Meng, T.L. and M. Khushi, Reinforcement learning in financial markets. Data, 2019. 4(3): p. 110.

[18] Li, Y., P. Ni, and V. Chang, Application of deep reinforcement learning in stock trading strategies and stock forecasting. Computing, 2020. 102(6): p. 1305-1322.

[19] Carta, S., et al., Multi-DQN: An ensemble of Deep Q-learning agents for stock market forecasting. Expert systems with applications, 2021. 164: p. 113820.

[20] Shi, Y., et al., Stock trading rule discovery with double deep Q-network. Applied Soft Computing, 2021. 107: p. 107320.

[21] Théate, T. and D. Ernst, An application of deep reinforcement learning to algorithmic trading. Expert Systems with Applications, 2021. 173: p. 114632.

[22] Sun, T., D. Huang, and J. Yu, Market Making Strategy Optimization via Deep Reinforcement Learning. IEEE Access, 2022.

[23] Jiang, W., Applications of deep learning in stock market prediction: recent progress. Expert Systems with Applications, 2021. 184: p. 115537.

[24] Amihud, Y., Illiquidity and stock returns: cross-section and time-series effects. Journal of financial markets, 2002. 5(1): p. 31-56.

[25] Jansen, S., Machine Learning for Algorithmic Trading: Predictive Models to Extract Signals from Market and Alternative Data for Systematic Trading Strategies with Python. 2020: Packt Publishing Limited.

[26] Mehrabanpour, M., A. Azar, and M. Shahrami Babkan, Stock price forecasting by presenting a hybrid model using principal component analysis and rough set theory. Modern Research in Decision Making, 2022. 7(2): p. 137-167.

[27] Minh, D.L., et al., Deep learning approach for short-term stock trends prediction based on two-stream gated recurrent unit network. Ieee Access, 2018. 6: p. 55392-55404.

[28] Hessel, M., et al. Rainbow: Combining improvements in deep reinforcement learning. in Thirty-second AAAI conference on artificial intelligence. 2018.

[29] Index. 1400; Available from: https://rahavard365.com/.