search for


A study on KOSPI 200 direction forecasting using XGBoost model
Journal of the Korean Data & Information Science Society 2019;30:655-69
Published online May 31, 2019;
© 2019 Korean Data and Information Science Society.

Dae Woo Hah1 · Young Min Kim2 · Jae Joon Ahn3

13Department of Information and Statistics, Yonsei University, 2Department of Bigdata Engineering, Soonchunhyang University
Correspondence to: Associate professor, Department of Information & Statistics, Yonsei University, Wonju 26493, Korea. E-mail:
This work was supported by the Yonsei University Wonju Campus Future-Leading Research Initiative of 2017 (2017-52-0072).
Received April 21, 2019; Revised May 16, 2019; Accepted May 16, 2019.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
The stock market is a representative market representing the capitalist economic system and performs important economic functions in the financial market. The stock market also plays a role of acquiring capital for both individual and corporate investors. Accurately predicting the stock prices remains an important research task. In recent years, studies on stock price prediction using machine learning have progressed. In this study, we will use the XGBoost (extreme gradient boosting) model, which has been used in various fields recently and proved its excellence to predict stock price fluctuation. In order to demonstrate the superiority of the XGBoost model, we compared and analyzed the results of the LSTM (long-short term memory) neural network which showed good performance in the previous studies and the autoregressive model which is a conventional time series analysis technique. The empirical analysis shows that the XGBoost model is competitive in predicting stock price movements.
Keywords : Autoregressive model, LSTM neural network, machine learning, Stock price prediction, XGBoost model.