The Study Of HS300 Index Predition Based On Machine Learning

With the development of artificial intelligence,quantitative investment has emerged a new trend and began to change to intelligent investment.Machine learning and deep learning algorithms have been widely used in quantitative investment.Based on the theories related to the Random Forest model(Random Forest)and the XGBoost model(eXtreme Gradient Boosting),this paper studies the related problems of the prediction of the intraday price trend of financial securities.This paper uses the most representative Shanghai and Shenzhen 300 index as the empirical research object.Firstly,according to the historical trading data and fundamental data of the Shanghai and Shenzhen 300 index,and the trading data of the related financial securities,the characteristic engineering is carried out to extract the characteristics of the effective technical analysis index and the fundamental analysis index.The Empirical Mode Decomposition(EMD)algorithm which is commonly used in digital signal processing is introduced.The trend component and concussion component in intraday price series of the index are decomposed,and the trend and concussion are classified by comparing the energy value of the two components.Then the random forest model and the XGBoost model are set up respectively.Forecast the single day trend classification of index price.This paper first introduces the development process of quantitative investment,and then introduces the theoretical basis of Random Forest model and XGBoost model,and finally uses the Shanghai and Shenzhen 300 index data to carry out an empirical analysis.The empirical process mainly includes three steps:feature extraction,trend classification,and model parameter optimization.Feature extraction combined with the actual investment experience,more comprehensive selection of the Shanghai and Shenzhen 300 index and related trading varieties of technical and basic analysis indicators.Parameter optimization is based on grid search algorithm,and K cross validation is used to select the highest accuracy parameter combination.The empirical results show that the Random Forest and XGBoost model have an ideal prediction effect on the price trend prediction of Shanghai and Shenzhen 300 index,and the accuracy of the forecast is both above 70%.It is obvious that the machine learning method has some guiding significance for the quantitative investment decision.