Predict stock price trends in Stock Exchange of Thailand using Ensemble Model

Main Article Content

ปรเมษฐ์ ธันวานนท์
ชัยกร ยิ่งเสรี
วรพล พงษ์เพ็ชร
ธนภัทร ฆังคะจิตร3


- The investment in the stock market has been interesting for domestic and foreign investors. There are many analysts attempting to improve forecasting models to increase prediction accuracy and stock return. Mostly focus on the single classifier using stock prices data as the main factor. However, the accuracy and stock return of the prediction were low. Thus, the purpose of this study attempt to increase prediction accuracy and stock return. Four factors were considered: stock prices data, indicator data, holding days and indicator days (the days that used to calculate indicators). In addition, 4 techniques of machine learning and ensemble model were used to forecast the trends of stock price by using SET’s information from January 2011 to December 2016. The results show that ensemble model in term of weight form can increase the prediction accuracy by 5% - 14% and increase the stock return by 1% - 3%. Moreover, the factors of indicators data and holding days are important factors to improve efficiency of prediction and stock return. KEY WORDS

Article Details

How to Cite
ธันวานนท์ ป., ยิ่งเสรี ช., พงษ์เพ็ชร ว., and ฆังคะจิตร3 ธ., “Predict stock price trends in Stock Exchange of Thailand using Ensemble Model”, JIST, vol. 7, no. 1, pp. 12–21, Jun. 2017.
Research Article: Soft Computing (Detail in Scope of Journal)


1. A.C. Charu, Outlier Analysis, 2nd ed. New York: IBM T.J. Watson Research Center Yorktown Heights, 2016.

2. B. Narayanon and M. Govindarajan, ”Prediction on Stock Market using Ensemble Model”, International Journal of Computer Applications, vol. 128, no. 1, pp. 18-21, Oct 2015.

3. B.V. Dasarathy, Nearest Neighbor Pattern Classification Techniques. California: IEEE Computer Society Press, 1991.

4. G. Shakhnarovich, T. Darrel and P. Indyk, Nearest Neighbor Methods in Learning and Vision, MIT Press, 2006.

5. H. Jiawei and K. Micheline, Data Mining Concepts and Techniques, 2nd ed. USA: Morgan Kaufmann, 2006.

6. J.T. Barkoulas and C.F. Baum, “Long-term dependence in stock returns”, Economics Letter, pp.253-259, Oct. 1996.

7. L. Khaidem, S. Saha and S.R. Dey, “Predicting the direction of stock market prices using random forest”, Applied Mathematical Finance, vol. 0, no. 0, pp. 1-20, May. 2015.

8. P. Sanguansat, Machine Learning. Thailand: Panyapiwat Instutute of Management, 2015.

9. P. Sutheebanjard and W. Premchaiswadi. “Stock Exchange of Thailand Index prediction using Back Propagation Neural Network”, Second International Conference on Computer and Network Technology, pp. 377-380, Apr. 2010.

10. S. Chaigusin, C. Chirathamjaree and J. Clayden. “The Use of Nueral Networks in the Prediction of the Stcok Exchange of Thailand (SET) Index”, International Conference on Computational Intelligence for Modeling Control & Automation, pp.670-673, Dec. 2008.

11. S. Nootyaskool, M. Attasiripanya and W. Prasidtichaikul. “Stock Forecasting with Candlestick Chart and Hidden Markov Models”, National Conference on Computing and Information Technology, 2014.

12. W. Huang, Y. Nakamori and S.Y. Wang, “Forecasting stock market movement direction with support vector machine”, Computers & Operations Research, vol. 32, no.10, pp. 2513-2522, Oct. 2005.

13. W. Qiang, C. Yuehui and L. Zhen, “Ensemble model of Intelligent Paradigms for stock Market Forecasting”, Knowledge Discovery and Data Mining, pp. 205-208, Jan. 2008.

14. Z. H. Zhou, Ensemble Methods Foundations and Algorithms, New York: CRC Press, 2012.