Abstract
The predominant method of developing trading strategies is technical analysis on historical market data. Other financial analysts monitor the public activity towards cryptocurrencies, in order to forecast upcoming trends in the market. Until now, the best cryptocurrency trading models rely solely on one of the two methodologies and attempt to maximize their profits, while disregarding the trading risk. In this paper, we present a new machine learning approach, named TraderNet-CR, which is based on deep reinforcement learning. TraderNet-CR combines both methodologies in order to detect profitable round trips in the cryptocurrency market and maximize a trader’s profits. Additionally, we have added an extension method, named N-Consecutive Actions, which examines the model’s previous actions, before suggesting a new action. This method is complementary to the model’s training and can be fruitfully combined, in order to further decrease the trading risk. Our experiments show that our model can properly forecast profitable round trips, despite high market commission fees.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
A DRL agent utilizes a deep learning model in order to learn to behave optimally in its environment.
- 2.
A round trip is a pair of two opposite orders placed one after the other (buy-sell or sell-buy), that aims to take advantage of price differences in order to produce profit.
- 3.
The sparse reward problem happens when an environment rarely produces a reward. This usually slows down the training process of a DRL agent [15].
- 4.
A signal is called bullish when the close price begins to rise. On the other hand, a signal is called bearish when the close price starts to drop.
- 5.
- 6.
OHLCV datasets consist of five columns: Open, High, Low, Close, Volume of a market at a specific time.
- 7.
References
Low, R.K.Y., Tan, E.: The Role of Analyst Forecasts in the Momentum Effect. Wiley Trading (2006)
Baiynd, A.M.: The Trading Book: A Complete Solution to Mastering Technical Systems and Trading Psychology. McGraw-Hill (2011)
Brown, R.G.: Smoothing, Forecasting and Prediciton of Time Series. Dover Publications (1963)
Brown, R.G.: New Concepts in Technical Trading Systems. Trend Research (1978)
Brown, R.G.: Technical Analysis Power Tools for Active Investors. Financial Times Prentice Hall (2005)
Gerstein, M.: Evaluation of the chaikin power gauge stock rating system. Chaikin Analytics (2013)
Granville, J.E.: Granville’s New Key to Stock Market Profits. Papamoa Press (2018)
Jiang, Z., Xu, D., Liang, J.: A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem, pp. 1–31 (2017). http://arxiv.org/abs/1706.10059
Livieris, I.E., Stavroyiannis, S., Iliadis, L., Pintelas, P.: Smoothing and stationarity enforcement framework for deep learning time-series forecasting. Neural Comput. Appl. 33(20), 14021–14035 (2021). https://doi.org/10.1007/s00521-021-06043-1
Low, R.K.Y., Tan, E.: The role of analyst forecasts in the momentum effect. Int. Rev. Finan. Anal. 9 (2016)
Lucarelli, G., Borrotti, M.: A deep q-learning portfolio management framework for the cryptocurrency market. Neural Comput. Appl. 32(23), 17229–17244 (2020)
Mudassir, M., Bennbaia, S. Unal, D.H.M.: Time-series forecasting of bitcoin prices using high-dimensional features: a machine learning approach. Neural Computing and Applications (2020)
Mulloy, P.: Technical Analysis of Stocks and Commodities 40(1) (1982)
Murphy, J.J.: Technical analysis of the financial markets: a comprehensive guide to trading methods and applications. Penguin (1999)
Noel, A.D., van Hoof, C., Millidge, B.: Online reinforcement learning with sparse rewards through an active inference capsule (2021)
Sattarov, O., Muminov, A., Lee, C.W., Kang, H.K., Oh, R., Ahn, J., Oh, H.J., Jeon, H.S.: Recommending cryptocurrency trading points with deep reinforcement learning approach. Appl. Sci. 10(4), 1506 (2020)
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., Klimov, O.: Proximal policy optimization algorithms (2017)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix: A
Appendix: A
A typical PPO Agent architecture
The TraderNet-CR actor-critic network architecture
Rights and permissions
Copyright information
© 2022 IFIP International Federation for Information Processing
About this paper
Cite this paper
Kochliaridis, V., Kouloumpris, E., Vlahavas, I. (2022). TraderNet-CR: Cryptocurrency Trading with Deep Reinforcement Learning. In: Maglogiannis, I., Iliadis, L., Macintyre, J., Cortez, P. (eds) Artificial Intelligence Applications and Innovations. AIAI 2022. IFIP Advances in Information and Communication Technology, vol 646. Springer, Cham. https://doi.org/10.1007/978-3-031-08333-4_25
Download citation
DOI: https://doi.org/10.1007/978-3-031-08333-4_25
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-08332-7
Online ISBN: 978-3-031-08333-4
eBook Packages: Computer ScienceComputer Science (R0)
