Decision support for the stocks trading using MLP and data mining techniques

• Main Authors: Narissara Eiamkanitchat, Teerasak Moontui
• Format: Book Series
• Published: 2018
• Subjects: Engineering
• Online Access: https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=84959125602&origin=inward; http://cmuir.cmu.ac.th/jspui/handle/6653943832/55777
• Institution: Chiang Mai University

© Springer Science+Business Media Singapore 2016. The investment in the stock market to buy shares requires extensive information for decision making. The prudent investors require understanding the numerous of fundamental information and studying many technical factors for the appropriate stock screening. The data mining, which includes many of the computational intelligence techniques, is proper to apply to these data. This research focuses on the high performance stock selections using the fundamental analysis of individual stocks, which is reflected in the financial statements. Total ten criteria calculate from the stock financial statements report is proposed to use for fundamental analysis. The MLP neural network is used in the training process of the five-year historical stock dataset for classifying good return stocks, those likely to win the market in the future. The short historical prices of the good return stocks are analyzed by using technical factors to identify the buying or selling signal in the decision support process. From the experimental results the Exponential Moving Average (EMA) technique is the most favorable and selected to apply in our system. The simulated investors are trading in the Sock Exchange of Thailand (SET) using the information of the decision support system in this research. The real information about the stock prices are used to evaluate the performance of the propose system. The average returns of the ports, that follow the system suggestion, are increased from the starting budget and almost triple time higher that market yield. The results show that the developed system is capable to filtering the good return stock, and suggest the proper signal of the investor. The return from the combination of selected fundamentals and technical can generate interesting returns that beat the overall market in comparison to the same period.