Prediction of stock price direction using a hybrid GA-XGBoost algorithm with a three-stage feature engineering process

作者：

Highlights：

• Good performance with less computational cost than deep learning models.

• Better interpretability than deep learning models.

• Feature set expansion to gain the blessing of dimensionality.

• Optimal feature set to relieve the curse of dimensionality.

摘要

•Good performance with less computational cost than deep learning models.•Better interpretability than deep learning models.•Feature set expansion to gain the blessing of dimensionality.•Optimal feature set to relieve the curse of dimensionality.

论文关键词：Genetic algorithm,XGBoost feature selection,Technical indicators,Blessing of dimensionality,Curse of dimensionality,Feature set expansion,Optimal feature set

论文评审过程：Received 9 March 2021, Revised 18 July 2021, Accepted 2 August 2021, Available online 13 August 2021, Version of Record 19 August 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115716