Research On Multi-factor Stock Selection Strategy Based On Random Forest Algorithm

With the continuous development of quantitative investment theory and the deepening of quantitative strategy research,the types of quantitative investment strategy can be continuously enriched.However,in the practical application process,the multi-factor stock selection strategy still performs better.On the one hand,it focuses on the analysis of traditional factors in the multi-factor model and the exploration of innovation factors.On the other hand,with the continuous maturity of machine learning algorithm,combining it with multi-factor stock selection model has become a research direction of strategy optimization.Through the comparison with the history and scale of quantitative investment in the United States,it is not difficult to find that the application history of quantitative investment in China is relatively short and the scale is relatively small.Therefore,quantitative investment in China's stock market application prospects are very considerable,the future development space is larger.Based on the traditional multifactor stock selection strategy and the random forest algorithm,this paper constructs a multi-factor stock selection model.In this paper,on the basis of predecessors' research,choose the a-share market the stock of the same industry data for model building and back to test,through the data preprocessing,through correlation and significance test of factor selection,determine the tree using K to fold the crossover and the number of features,and the model to measure test,the model of earnings compared with Shanghai and CSI 300 index,found that portfolio returns is superior to control group,and the random forest algorithm is applied to many other factors model,and find that the strategy also obtained is higher than the benchmark results.It shows that this strategy has strong feasibility in China's stock market.