Stock Quantitative Transaction Based On Support Vector Machine Operational Research

In recent years,the development oriented quantitative investment of science and technology has been paid more and more attention.From the current situation of China's economic development,quantitative investment will have a good prospect for a long time in the future.It is precisely because of the great potential of quantitative investment that the quantitative investment will be made by many investors.However,because the scale and strategy of the quantitative investment in our country are single,the performance is not satisfactory,so it is necessary to study the new method of quantitative investment and model,and the rapid development of machine learning has also brought great opportunities for the development of quantitative investment.The basic idea of SVM is to transform the classification problem of low dimensional space into the classification problem of high dimensional space after the selection of the corresponding kernel function,and the linear model which maximizes the requirement interval of high dimensional space makes the data can be divided in the high dimensional space.When evaluating the SVM,this paper first establishes the error rate of the model,then points out the limitations of the other models in calculating the judgement rate,and proves that the error rate model proposed in this paper has a significant improvement on the other misjudgment rates.At the same time,the method of cross validation is introduced to find the rate of miscarriage of justice.Based on the classification technology of SVM in machine learning,this paper uses the classification index of stock as the basis of classification,and classifies the stock to predict the return rate of the stock at the same time.After the classification results are obtained,the outstanding stock and the bad stock construction combination are selected and the yield is finally greater than that of the stock.The average rate of return of the market.In this paper,the financial index data of more than 900 stocks are selected.After the principal component analysis,the characteristics of the variables are selected.The contribution rate of the variance is 97.54%.For the selected principal components,the classification technique of the support vector machine is used and the Gauss kernel function is improved.The classification machine with better classification results is obtained,and then the combination of dominant stock is obtained.On the forecast of stock returns,based on the three factor combination proposed by the predecessors,the combination model of the parameter improvement three factor is added,and the forecast of the yield is improved obviously than the income in the market.This shows that this paper,The proposed stock selection model based on support vector machine is available,which provides a good reference for the establishment of new quantitative stock selection and quantitative prediction in the future.