Research On Forecasting And Investment Strategy Of Shanghai-Shenzhen 300 Index Based On Support Vector Machine

The results show that the Hurst index of financial market returns is between 0.5 and 1,indicating that financial markets are sensitive to initial conditions,that is,financial markets have predictability.Since the Shanghai and Shenzhen 300 index contains 300 stocks with high market value and good liquidity,its changes are mainly affected by systemic risk,which is relatively easy to predict.On the other hand,changes in stock index prices can provide effective investment information for financial market participants.As a new machine learning method,SVM can effectively avoid over-fitting and "dimension disaster" when dealing with non-linear and inseparable financial data.Moreover,its algorithm has strong learning performance,so it has great advantages in stock price prediction.Based on the classification principle of SVM and the bifurcation principle of decision tree and Stochastic Forest algorithm,this paper establishes a three-class forecasting model of ShanghaiShenzhen 300 index,that is,on the basis of judging the direction of the change of ShanghaiShenzhen 300 index,the magnitude of the change of Shanghai-Shenzhen 300 index is also judged accordingly.In detail,normalizing the initial feature and putting the value feature and the small value feature on the same magnitude can not only lay a good foundation for feature selection,but also narrow the parameter range of the model,improve the classification accuracy of the model,and improve the training speed of the model.In order to improve the accuracy of the prediction model,the kernel function is screened in the research process,and the parameters are optimized by 5-fold cross validation and grid search.The best kernel function and parameters are used to construct the final prediction model.Finally,the investment strategy is established according to the prediction results of the model,and the effect of the model is tested from the perspective of actual combat.