Optimized stock multi-factor model based on Machine Learning

• 1、China Three Gorges University,School of Computer and Information，YiChang 443002

Abstract：The purpose of this paper is to build an optimized stock multi-factor model based on machine learning to deal with A-share market style switching and stock picking issues to finally obtain excess returns. From the perspective of factor expression and machine learning algorithm, the paper studies the fluctuation law of A-share market stocks and obtains the maximum return of excess returns. First, a factor analysis model was constructed to screen out seven optimal factors. Then a stochastic forest model based on machine learning is constructed, and the stock fluctuations of a certain period of time are measured through random forests. This paper uses the published data from China\'s A-share market from January 1, 2016 to September 30, 2018 to evaluate the performance of the algorithm. The experimental results show that the correct rate of backtesting is 83%. The average interest rate on earnings is about 1.57%.

Keywords： Computer application excess return random forest entropy risk Machine Learning algorithm