Abstract
The paper describes the relevance of machine learning methods, namely training with reinforcement, to the problems of predicting financial time series. An overview of existing applications based on machine learning in the issues of financial market forecasting is presented. The reasons for the popularity of my research topic are highlighted both from the scientific (increasing the number of publications on issues relevant to the research topic over the past five years) and from the practical side. The analysis of scientific works, the subject and purpose of which are related to the issues and objectives of my research and their features are presented. The main problems associated with the problem of predicting stochastic time series are identified. According to the analysis, the purpose of work is defined, and also the list of tasks for the achievement of the set goal is made. The article is devoted to studying the use of the ensemble of neuro-learning networks with the strengthening of the securities trading market. The practical significance of the work is to use the model of efficient distribution of investments in the market. This paper will explore a set of models that use in-depth consolidated learning to explore trading strategies to maximize return on investment. The potential of using acting-critical models as an ensemble has been studied. Models such as Proximal Policy Optimizer (PPO), Advantage Actor-Critic (A2C) and Deep Determinist Police Gradient (DDPG) were used to teach trading strategy. To adapt the model to different situations, analyzes are analyzed according to three algorithms: the Dow Jones average and a portfolio that minimizes fluctuations in the Charpy ratio by balancing risk and return. The article compares ensembles by the method of fixing deep neural networks. To optimize the balance of risk and profit, the indicators of the ensemble model are analyzed. The ensemble and three-component models that apply well to market collapse conditions are considered. The models have learned to use the turbulence index for early stock sales to minimize losses during a stock collapse. The turbulence index threshold is demonstrated using ensemble models to regulate risk avoidance.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Agarwal, A., Kakade, S.M., Lee, J.D., Mahajan, G.: On the theory of policy gradient methods: optimality, approximation, and distribution shift (2019). https://doi.org/10.48550/ARXIV.1908.00261, https://arxiv.org/abs/1908.00261
Agarwal, A., Kakade, S.M., Lee, J.D., Mahajan, G.: Optimality and approximation with policy gradient methods in Markov decision processes. In: Proceedings of Thirty Third Conference on Learning Theory, pp. 64–66. PMLR (2020), https://proceedings.mlr.press/v125/agarwal20a.html
Alberg, D., Shalit, H., Yosef, R.: Estimating stock market volatility using asymmetric GARCH models. Appl. Finan. Econ. 18(15), 1201–1208 (2008). https://doi.org/10.1080/09603100701604225, http://www.tandfonline.com/doi/full/10.1080/09603100701604225
Assran, M., Romoff, J., Ballas, N., Pineau, J., Rabbat, M.: Gossip-based actor-learner architectures for deep reinforcement learning (2019). https://doi.org/10.48550/ARXIV.1906.04585, https://arxiv.org/abs/1906.04585
Babaeizadeh, M., Frosio, I., Tyree, S., Clemons, J., Kautz, J.: Reinforcement learning through asynchronous advantage actor-critic on a GPU (2016). https://doi.org/10.48550/ARXIV.1611.06256, https://arxiv.org/abs/1611.06256
Bhatnagar, S., Sutton, R.S., Ghavamzadeh, M., Lee, M.: Natural actor–critic algorithms. Automatica 45(11), 2471–2482 (2009). https://doi.org/10.1016/j.automatica.2009.07.008, https://linkinghub.elsevier.com/retrieve/pii/S0005109809003549
Boyko, N.: Application of mathematical models for improvement of “cloud" data processes organization. Math. Model. Comput. Sci. J. Comput. Prob. Electrotech. 3(2), 111–119 (2016). https://doi.org/10.23939/mmc2016.02.111
Boyko, N., Kmetyk-Podubinska, K., Andrusiak, I.: Application of ensemble methods of strengthening in search of legal information. In: Babichev, S., Lytvynenko, V. (eds.) ISDMCI 2021. LNDECT, vol. 77, pp. 188–200. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-82014-5_13
Busoniu, L., Bruin, T., Tolic, D., Kober, J., Palunko, I.: Reinforcement learning for control: performance, stability, and deep approximators. Ann. Rev. Control (2018). https://doi.org/10.1016/j.arcontrol.2018.09.005
Chague, F., De-Losso, R., Giovannetti, B.: Day trading for a living? SSRN Electron. J. (2019). https://doi.org/10.2139/ssrn.3423101
Chong, T., Ng, W.K., Liew, V.: Revisiting the performance of MACD and RSI oscillators. J. Risk Finan. Manag. 7(1), 1–12 (2014). https://doi.org/10.3390/jrfm7010001, http://www.mdpi.com/1911-8074/7/1/1
Čermák, M., Malec, K., Maitah, M.: Price volatility modelling - wheat: GARCH model application. Agris On-line Pap. Econ. Inf. 09(04), 15–24 (2017). https://doi.org/10.7160/aol.2017.090402
Gurrib, I.: Performance of the average directional index as a market timing tool for the most actively traded USD based currency pairs. Banks Bank Syst. 13(3), 58–70 (2018). https://doi.org/10.21511/bbs.13(3).2018.06
Kumar, H., Koppel, A., Ribeiro, A.: On the sample complexity of actor-critic method for reinforcement learning with function approximation (2019). https://doi.org/10.48550/ARXIV.1910.08412, https://arxiv.org/abs/1910.08412
Li, H., Dagli, C.H., Enke, D.: Short-term stock market timing prediction under reinforcement learning schemes. In: 2007 IEEE International Symposium on Approximate Dynamic Programming and Reinforcement Learning, pp. 233–240 (2007). https://doi.org/10.1109/ADPRL.2007.368193
Moody, J., Saffell, M.: Learning to trade via direct reinforcement. IEEE Trans. Neural Netw. 12, 875–889 (2001). https://doi.org/10.1109/72.935097
Neuneier, R.: Optimal asset allocation using adaptive dynamic programming. In: Conference on Neural Information Processing Systems, vol. 8, pp. 952–958. MIT Press (1996). https://doi.org/10.5555/2998828.2998962
Qiu, S., Yang, Z., Ye, J., Wang, Z.: On finite-time convergence of actor-critic algorithm. IEEE J. Sel. Areas Inf. Theory 2(2), 652–664 (2021). https://doi.org/10.1109/JSAIT.2021.3078754, https://ieeexplore.ieee.org/document/9435807/
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., Klimov, O.: Proximal policy optimization algorithms (2017). https://doi.org/10.48550/ARXIV.1707.06347, https://arxiv.org/abs/1707.06347
Shen, J., Shafiq, M.O.: Short-term stock market price trend prediction using a comprehensive deep learning system. J. Big Data 7(1), 66 (2020). https://doi.org/10.1186/s40537-020-00333-6, https://journalofbigdata.springeropen.com/articles/10.1186/s40537-020-00333-6
Sun, S., Wang, R., An, B.: Reinforcement learning for quantitative trading. arXiv:2109.13851 [cs, q-fin] (2021), http://arxiv.org/abs/2109.13851, arXiv: 2109.13851
Thomas, G.F.: Reinforcement learning in financial markets - a survey. Technical Report 12/2018, Friedrich-Alexander University Erlangen-Nuremberg, Institute for Economics (2018). https://www.econstor.eu/bitstream/10419/183139/1/1032172355.pdf
Xu, T., Wang, Z., Liang, Y.: Improving sample complexity bounds for (natural) actor-critic algorithms (2020). https://doi.org/10.48550/ARXIV.2004.12956, https://arxiv.org/abs/2004.12956
Xu, T., Wang, Z., Zhou, Y., Liang, Y.: Reanalysis of variance reduced temporal difference learning (2020). https://doi.org/10.48550/ARXIV.2001.01898, https://arxiv.org/abs/2001.01898
Yang, Z., Zhang, K., Hong, M., Basar, T.: A finite sample analysis of the actor-critic algorithm. In: 2018 IEEE Conference on Decision and Control (CDC), pp. 2759–2764. IEEE, Miami Beach (2018). https://doi.org/10.1109/CDC.2018.8619440, https://ieeexplore.ieee.org/document/8619440/
Zhang, Y., Yang, X.: Online portfolio selection strategy based on combining experts’ advice. Comput. Econ. 50(1), 141–159 (2017). https://doi.org/10.1007/s10614-016-9585-0, https://doi.org/10.1007/s10614-016-9585-0
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Boyko, N. (2023). The Comprehensive Model of Using In-Depth Consolidated Multimodal Learning to Study Trading Strategies in the Securities Market. In: Babichev, S., Lytvynenko, V. (eds) Lecture Notes in Data Engineering, Computational Intelligence, and Decision Making. ISDMCI 2022. Lecture Notes on Data Engineering and Communications Technologies, vol 149. Springer, Cham. https://doi.org/10.1007/978-3-031-16203-9_8
Download citation
DOI: https://doi.org/10.1007/978-3-031-16203-9_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-16202-2
Online ISBN: 978-3-031-16203-9
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)