Prediction of stock price direction using XGBoost algorithm

Abstract

This paper predicts the Vnindex’s direction using technical indicators and Extreme gradient
boosting. In addition, the study find out the important of prediction features. The data used used 5406
trading days’ worth of data regarding the Vnindex to predict the stock index’s up or down movement direction
on the next day form 09/10/2000 to 28/ 12/2022 at cophieu68.vn. The technical indicators confirm the
forecast performance with over 80 percentage accuracy and the volume is the most important feature to
predtict the trend of Vnindex indicator.