Forecasting exchange rate has been regarded as one of the most challenging task of modern financial time series analyses for many years. This work has goal whether using Artificial Neural Network (ANN) in high frequency exchange rate is applicable or not. Then neural network predicting performances are compared with Support Vector Machine (SVM) and Long Short Term Memory Neural Network (LSTM). Exchange rates used in experiments are EURO/USD (euro, dollar), USD/JPY (dollar, Japanese, yen), GBP/USD (pound, dollar) USD/CHF (dollar, Swiss franc) and 1 min, 5 min, 10 min, 30 min, time frames have been chosen. Forecasting instrument is Artificial Neural Networks as a machine learning method

Abstract I
Table of Contents II
List of Figures IV
List of Tables V
Chapter 1 Machine learning techniques 1
Introduction 1
Literature Review 4
1.1 Support vector machine 6
1.2 Support vector regression 8
1.3 Application of support vector machine 9
1.4 Artificial neural network 9
1.5 Cost function 11
1.6 Gradient Descent 11
1.7 Backpropagation algorithm 12
1.8 Activation function 12
1.9 Long Short Term Memory 14
1.10 Mani concept 15
Chapter 2 Methodology 19
2.1 Our dataset 19
2.2 Basic statistics on data 19
2.3 Forecasting methodology 22
2.4 Data pre-processing 22
2.5 SVM model 23
2.6 ANN model 23
2.7 LSTM model 25
2.8 One-step-ahead predictions 25
2.9 Training and validation set 25
2.10 Forecasting evaluation and accuracy measures 26
2.11 Statistical Indicators of RMSE 26
2.12 An economic indicator: Theil’ U rate 27
2.13 Direction Accuracy 27
Chapter 3 Implementation 29
3.1 Support vector machine 29
3.2 Artificial neural network 29
3.3 Long Short Term Memory 30
Chapter 4 Out-of-sample forecasting performance results 31
4.1 SVM result 31
4.2 ANN result 32
4.3 LSTM result 33
4.4 Comparison of 3 model 34
Chapter 5 Conclusions 35
5.1 Conclusions 35
5.2 Research limitation 35
5.3 Suggestions for future work 36
Reference 37
Appendix I – Other SVM paper summary 40
Appendix II – Other ANN paper summary 42

Alamili, M., (2011). Exchange Rate Prediction using Support Vector Machines: A comparison with Artificial Neural Networks (Master‘s Thesis). Retrieved from
https://pdfs.semanticscholar.org/a634/ad956231d40026def9a00ecb339500f75fff.pdf.
Bao,W., Yue, J., Rao, Y., 2017. A deep learning framework for financial time series using using stacked autoencoders and long-short term memory. PLoS ONE 12(7): e0180944. https://doi.org/10.1371/journal.pone.0180944.
Cao, Z.D., Pang, S.L., Bai, Y.H., 2005. Forecasting exchange rate using Support Vector Machine. Retrieved from 10.1109/ICMLC.2005.1527538.
Choudhry, T., McGroarty, F., Peng, K., Wang, S., 2012. High frequency exchange rate prediction with artificial neural network. Volume 19, Issue 3 July/September 2012 Pages 170-178.
Fletcher, T., 2010. Machine Learning for Financial Market Prediction (PhD Thesis). Retrieved from http://discovery.ucl.ac.uk/1338146/1/1338146.pdf.
Galeshchuk, S., 2016. Neural networks performance in exchange rate prediction.
Journal of neuro computing volume 172 issue, 2016 pages 446-452.
Galeshchuk, S., Mukherjee, S., 2017. Deep Networks for Predicting Direction of Change in Foreign Exchange Rates. Volume24, Issue4 October/December 2017 Pages 100-110.
Hansson, M., 2017. On stock return prediction with LSTM networks (Master‘s Thesis). Retrieved from http://lup.lub.lu.se/luur/download?func=downloadFile&recordOId=8911069&fileOId=89110
Huang, W., Nakamori, Y., Wang, Y., 2005. Forecasting stock market movement direction with support vector machine, Computers & Operations Research, Volume 32, Issue 10, Pages 2513-2522.
Kamruzzman, J., Sarker, R., 2004. Application of support vector machine for forex monitoring: Empirical Methods in Natural Language Processing - Volume 10, pages 79—86.
Kercheval, A., Zhang, Y., 2013. Modeling high-frequency limit order book dynamics with
with support vector machines (Master‘s Thesis). Retrieved from
https://www.math.fsu.edu/~aluffi/archive/paper462.pdf.
Krollner, B., Vanstone, B., and Finnie, G., 2010. Financial time series forecasting with
machine learning techniques: A survey.
Palikuca, A., Seidl, T., (2016). Predicting High Frequency Exchange Rates using Machine Learning (Master‘s Thesis). Retrieved from
http://www.diva-portal.org/smash/get/diva2:932585/FULLTEXT02.
Pandaa, C., Narasimhan, V., 2006. Forecasting exchange rate better with artificial neural network: Journal of Policy Modeling 29(2):227-236 • March 2007. 
Qu, H., and Zhang, Y., 2015. A New Kernel of Support Vector Regression for Forecasting H High-Frequency Stock Returns: Mathematical Problems in Engineering
Volume 2016, Article ID 4907654, 9 pages.
Singh, R., and Balasundaram, S., 2007. Application of extreme learning machine method for

time series analysis. International Journal of Intelligent Technology, 2(4):256–262.
Stokes, A., Zaid, A., 2012. Forecasting foreign exchange rates using artificial neural networks : a trader's approach. Int. J. Monetary Economics and Finance, Volume.5, No. 4, 2012.
Tenti, P., 1996. Forecasting exchange rate using recurrent neural network.
Applied artificial intelligence 10:567-581, 1996.
Trebal, A., (2009). High frequency time series forecasting with application to stock market (Master‘s Thesis). Retrieved from
https://www.politesi.polimi.it/bitstream/10589/5062/1/TESI_arnaud_trebaol.pdf.
Walczak, S., 2001. An empirical analysis of data requirements for financial forecasting with
neural networks. Journal of management information systems, 17(4):203–222.
Wu, B., 1995. Model-free forecasting for nonlinear time series with application to exchange rates: Volume 19, Issue 4, Pages 433-459.
Yuan, Y., 2012. Forecasting movement of direction exchange rate with polynomial smooth support vector machine: Volume 57, Issues 3–4, February 2013, Pages 932-944.
Zankova, E., (2016). High frequency financial time series prediction: machine learning approach (Master‘s Thesis). Retrieved from
https://munin.uit.no/bitstream/handle/10037/9255/thesis.pdf?sequence=2.
Zhang, G., Michael, Y., 1997. Neural Network Forecasting of the British Pound/US Dollar Exchange Rate: Volume 46 Issue 3, December 2017 Pages 1095-1119.
Zhuge, Q., Xu, L., and Zhang, G., 2017. LSTM Neural Network with Emotional Analysis for Prediction of Stock Price. Engineering Letters, 25:2, EL_25_2_09.
Understanding LSTM networks
http://colah.github.io/posts/2015-08-Understanding-LSTMs/.