Research on Quantitative Stock Selection Method Based on Random Forest

Stock investment has always been a hot issue in the financial and investment fields. The research of stock quantitative investment method based on modern data mining method and machine learning technology is a brand new field. Random forest is a prediction method integrating multiple decision trees. This paper studies the application of random forest in the quantitative stock selection of stocks, selects the annual report data of China and Shenzhen 300 constituent stocks from 2014 to 2018, and compares the prediction of stock investment returns by using decision tree and random forest method respectively. The applicability and prediction performance of the random forest method are verified, and the importance and interrelationship of annual report indicators are preliminary explored.