Study On Quantitative Timing Strategy Of Shanghai And Shenzhen 300 Index With SVM

With the continuous progress of global information science and technology,more and more researchers begin to study the capital market by means of machine learning.There are many problems in using the traditional methods to analyze the capital market indicators,such as requiring investors to have certain investment experience and ability,the investment decision-making is greatly affected by the investor's mood fluctuation,and many factors of capital market disturbance are difficult to quantify.In recent years,with the characteristics of independence,timeliness and systematicness,quantitative investment has been gradually popularized from institutional investors to more and more individual investors.In this paper,support vector machine(SVM)is used to build a quantitative timing model to analyze the data of Shanghai and Shenzhen 300 index.The key indexes such as characteristic indexes and model parameters in SVM quantitative timing model are studied.The feasibility and validity of the model are proved by experimental results and theoretical analysis.Specifically,the main research contents of this paper are as follows:1.In order to solve the shortcomings of technical analysis and fundamental analysis in traditional investment methods,this paper chooses Shanghai and Shenzhen300 index as the research object and proposes a quantitative timing strategy based on SVM theory.Through the establishment of the model to train the historical data of the market,and compare the predicted results with the actual market.Finally,the paper introduces correlation analysis index to evaluate the model,and tests the feasibility of the model through empirical analysis.2.In order to eliminate the noise in the Shanghai and Shenzhen 300 index market data and retain the main information of the data to the maximum extent,13 representative market indicators are selected as sample characteristic indicators in this paper.Through the discretization of the characteristic index data,the original continuous market data is transformed into the data with binary characteristics,furtherreducing the impact of market noise on the SVM quantitative timing strategy model.3.In order to improve the prediction accuracy,the grid search and cross validation are used to optimize the penalty factor C and the insensitive factor ? in SVM kernel function.Compared with other research methods,the model established in this paper can optimize and update the kernel parameters before each training and prediction.Through the research and experimental results of this paper,it shows that the SVM quantitative timing strategy model is feasible and accurate to predict the trend of Shanghai and Shenzhen 300 index.To some extent,the model can obtain excess return,which has certain reference value and guiding significance for investors' decision-making.