近年來隨著機器學習的進步，圍棋AI，AlphaGo擊倒柯傑(世界圍棋冠軍)，研究深度學習的學者也越來越多，而一貫走在前沿的金融界也越來越關注深度學習在金融交易中的應用。

本論文建制了一個智能交易系統，包含長短期記憶遞迴類神經網路(long short-term memory recurrent neural network)的預測模型用以預測外匯市場及設計一個外匯自動交易程式，該程式以MetaTrader 5為開發平台。該系統預測歐幣兌美元之每分鐘漲跌方向準確率為52.1%，在本金為10000美金下，我們實際進行近三年的操盤，在mt5平台上皆具有不錯的獲利。

In recent years, with the advancement of machine learning, Go AI, AlphaGo have knocked Ke Jie (World Go Championship), and more and more scholars have studied deep learning. The financial community is also increasingly concerned about the application of deep learning in financial transactions.

This paper constructs an intelligent trading system, including a long-term memory recursive neural network prediction model, used to forecast the foreign exchange market and design a forex automatic trading program, the program developed by MetaTrader 5 platform. The system predicts that the accuracy of the EURUSD trend is 52.1%. Under the balance of 10,000 US dollars, we use our strategy to test the true data in past three years with deep learning, and we get the positive profit.

Contents................................................i
List of Figures.........................................iii
List of tables..........................................v
Chapter 1 Introduction................................1
1.1 Motivation..........................................2
1.2 Organization................................3
Chapter 2 Background..................................4
2.1 Efficient Market theory.....................4
2.2 Neural Networks.............................6
2.2.1 Recurrent Neural Networks.........6
2.2.2 LSTM..............................9
2.3 MetaTrader 5................................12
2.3.1 Overview..........................12
2.3.2 MQL5..............................12
2.3.3 Martingale Strategy.......13
Chapter 3 Relative Work...............................15
Chapter 4 Implement...................................19
4.1 Overview............................................19
4.2 Data Selection..............................20
4.3 LSTM Model..........................................21
4.4 Our EA......................................24
Chapter 5 Experimental Result.........................28
5.1 Model Predictions of Trends.........................28
5.2 EA Performance..............................30
5.3 Discuss.............................................33
Chapter 6 Conclusion..................................34
6.1 Conclusion..........................................34
6.2 Future Work......................................35
Bibliography............................................36
Appendix................................................39

[1]Ashwin Siripurapu, Convolutional Networks for Stock Trading, Stanford University Department of Computer Science. http://cs231n.stanford.edu/reports/2015/pdfs/ashwin_final_paper.pdf, 2015.

[2]Yun-Cheng Tsai, Jun-Hao Chen, Jun-Jie Wang, Predict Forex Trend via Convolutional Neural Networks, https://arxiv.org/pdf/1801.03018.pdf

[3]L. Di Persio, O. Honchar, Artificial neural networks approach to the forecast of stock market price movements, International Journal of Economics and Management Systems, 1:158–162, 2016.

[4]Understanding LSTM Networks, http://colah.github.io/posts/2015-08-Understanding-LSTMs/.

[5]Gidofalvi, G., & Elkan, C., Using news articles to predict stock price movements. Department of Computer Science and Engineering, University of California, San Diego, 2006.

[6]Li, Q., Wang, T., Li, P., Liu, L., Gong, Q., & Chen, Y., The effect of news and public mood on stock movements. Information Sciences, 278, 826-840,2014.

[7]Eugene F. Fama, Efficient Capital Markets: A Review of Theory and Empirical Work, American Finance Association, 1970

[8]Lester Ingber, Statistical Mechanics of Nonlinear Nonequilibrium Financial Markets: Applications to Optimized Trading, Mathematical and Computer Modelling, 23(7):101-121, 1996


[9]S. Taylor., Modelling financial time series, 1986

[10]S. Hochreiter, J. Schmidhuber, Long short-term memory, Neural Computation 9(8):1735-1780, 1997

[11]Matthew Dixon, Diego Klabjan, and Jin Hoon Bang, Classification-based Financial Markets Prediction using Deep Neural Networks, Available at SSRN: https://ssrn.com/abstract=2756331, 2016

[12]L. Takeuchi, Yu-Ying (Albert) Lee, Applying Deep Learning to Enhance Momentum Trading Strategies in Stock, 2013

[13]Gilberto Batres-Estrada, Deep learning for multivariate financial time series, 2015

[14]Chengzhang Zhu Jianping Yin, Qian LI, A Stock Decision Support System based on DBNs, Journal of Computational Information Systems 10(2): 883–893, 2014

[15]Samuel Rönnqvist, Peter Sarlin, Bank distress in the news: Describing events through deep learning, Neurocomputing, 264, 2017

[16]Ralph Fehrer, Stefan Feuerriegel, Improving Decision Analytics with Deep Learning: The Case of Financial Disclosures, 2015

[17]Xiao Ding, Yue Zhang, Ting Liu, Junwen Duan, Deep Learning for Event-Driven Stock Prediction, Artificial Intelligence IJCAI, 2015

[18]J. B. Heaton, N. G. Polson, J. H. Witte, Deep Portfolio Theory, 2016

[19]Abhijit Sharang, Chetan Rao, Using machine learning for medium frequency derivative portfolio trading, 2015

[20]Geoffrey E. Hinton, Simon Osindero, A fast learning algorithm for deep belief nets, Neural Computation 18 (7) 1527 - 1554, 2006