出版社：Canadian Research & Development Center of Sciences and Cultures
其他摘要：This paper proposed a hybrid model based on triangular fuzzy information granulation and SVM to predict the trend and fluctuation range of the price of stock index futures. Firstly, the original data is processed by triangular fuzzy information granulation. Then, the cross validation method is used to obtain the parameters of SVM. Also some significant factors concerning investors sentiment are considered to improve the forecast accuracy of the hybrid model. At last, the hybrid model is used to perform empirically study based on HS300 stock index futures data after fuzzy information granulation. The empirical analysis showed that the hybrid model owns the better preformation for the prediction of the change trend and range of the price of stock index futures. After concerning investors sentiment, the accuracy of prediction is also improved effectively.
关键词：Stock index future;Fuzzy information granulation;Support Vector Machine (SVM);Investors sentiment
Loading...
