Event Driven System Stock Price Forecast Model

Event-driven strategy has already developed into and keeps abreast of industry configuration and asset allocation strategy research method.In the early mining and analyzing the possible causes of the abnormal stock price fluctuations event,based on the grasp the trading time can let investors get excess returns on the event-driven stock.SVM has been applied to predict the stock price more and more frequently in recent years due to its great capability to deal with the non-linear relation.This thesis predicts the price of certain stocks by using the support vector machine and the capital flow as the index of event driven stock.The predictions of this model is above 80% accurate based on empirical analysis.In this paper,the first chapter is introduction,mainly elaborated the writing background of this article and the innovation points,as well as the domestic and foreign research on related topics.The second chapter is the precondition of the paper,mainly introducing the two of the usual indicator in security market,price and price index.This chapter also summarizes several common methods to predict the price of stock.The third chapter is about the event-driven strategy theory analysis,explained the reason why event affect the stock price,and the two event-driven strategy fund case.The fourth chapter includes the mathematical principle of support vector machine(SVM)and how it is to predict stock price.The fifth chapter is the key of the article,to predict stock price model is established,and through the empirical analysis shows that the accuracy of the model.The sixth chapter put forward the advantages and disadvantages in this paper.