Stock Prediction Research Based On Weighted Support Vector Machine Algorithm Using Grey Correlation Degree

The fast-paced progress of artificial intelligence and big data analytics is propelling the advancemen,the processing and analysis of massive financial data have become more efficient and effective,and the ability to mine text language features has been constantly improved.To improve the accuracy of stock price prediction,this paper improves the Support Vector Machine Regression algorithm by using gray correlation analysis.Also,the research on the text emotional tendency directly relating to the company’s operating performance and stock price in the annual report of listed companies has improved the accuracy of stock prediction.This paper first divides the factors that affect the stock price trend into behavioral factors and technical factors.The behavioral factors are mainly investor sentiment factors,and the technical factors mainly include daily closing data and the CSI 300 Index.Then the relationship between the stock price and its influencing factors in multiple trading days is depicted by the method of Grey Relational Analysis(GRA),and the relationship is converted into the characteristic weight of each influencing factor.The characteristic weight is used to modify the weight of the influencing factors in all trading days,and the modified stock trading data is predicted by Support Vector Machine Regression(SVR).Based on the technical indicators(MSE,MAE,SCC and DS),the forecast results of the revised and unmodified transaction data were compared,and it was found that the revised forecast was significantly improved.