Dynamic Weighting Multi Factor Stock Selection Strategy Based on XGboost Machine Learning Algorithm

Abstract

Tree boosting is a highly effective and widely used machine learning method. A dynamic weighting multi-factor stock selection strategy based on XGBoost model is constructed. XGboost machine learning method is used to predict the IC coefficients of factors. The results of back testing show that the performance of dynamic weighting strategy is superior to the equal weighting strategy and IC weighting strategy. The empirical results prove that XGBoost model is effective in predicting IC coefficients and the dynamic weighting based on XGBoost model can improve the performance of multi-factor stock selection strategy.