A Study Of Quantitative Timing Strategy Based On Deep Learning

As an important part of our market economy,the stock market is playing a more and more important role in the economic life and there exits more and more literatures on quantitative investment strategies in China.In this paper,HS300 as the research target,in order to improve the investment profitability and stability,based on deep learning of stock prediction model,and the results of the model combined with the corresponding timing trading strategies build the model of quantitative timing strategy based on deep learning.Because HS300 index has launched in China,as the research target,HS300 has practical significance in study of investment strategy.Using some traditional method to study the securities market,securities market not only need to meet more a series of stringent assumptions,but also it is difficult to obtain more satisfactory results.And nonlinear data mining technology can well explain the behavior of stock market price,therefore this paper,by using deep learning and support vector machine(SVM)of two kinds of nonlinear data mining technology to analyze our country behavior of stock market price,and comparing the results shows that deep learning in predicting the future of the price of the stock has a higher and more stable forecast accuracy.In view of this,this paper using deep learning,and connecting with the concrete data judges and classify the future of the stock price,the long-short operation is based on classification results.At the same time using support vector machine(SVM)to establish the prediction model of stock classification operation and according to the results carry out the long-short operation.Comparing deep study and the comparison of support vector machine(SVM),it is concluded that the classification of stock prediction based on the deep learning model has higher prediction accuracy,but also based on the deep study of quantitative timing strategy has higher and stronger stability returns and defensive ability.