Application Of The Securities Market Based On Data Mining Technology

Financial markets as the more common an important place for the collection and circulation of capital,is an important part of market economy,which function is that connecting investors and financing to collect the small money of the society to meet the needs of the focus on economic development through the relevant rules.At the same time,the financial market have a certain effect on the capital to regulate the market and maintaining the stability of the capital within a certain range better resolve the capital supply and demand contradiction to some extent.Therefore,the research on financial market investment analysis is the area of economic scholars keen.In numerous financial investment research model,the most common one is the traditional model based on time series,which is based on certain assumptions,but these hypothetical conditions consistent with the actual degree is low,which is the limitation of the model.In addition,the traditional model based on time series is a global model fixed structure,but due to the time series of financial market s tructure is unstable,which changes at any time,so using this model to describe is not very precise.With the continuous development of society,the information technology have gone into the financial industry,which promotes the vigorous development of the industry,at the same time,the financial market information is generated by the increased exponentially,so people also face a new problem,how to better and faster from vast amounts of financial information to obtain the information to their own needs.Data mining technology is to solve the problem to get useful information from vast amounts of information in the emerging technology,which greatly solves the confusion and brings a lot of convenience to people in the access to information.The paper,based on the existing problems which come from the existing FCM clustering model combined with the EMD-the SVR forecasting model algorithm,proposes based on the improved FCM clustering algorithm of the EMD-the SVR forecasting model.In the first place,the model does the cluster to data to fact the classification of the training data objects and calculation of the membership degree to reduce the number of input rules article to simplify the EMD-the SVR forecasting model finally.Applying the model respectively in Shenzhen and Shanghai stock index prediction to get the experimental results shows that the prediction model of prediction error is lower than used alone the EMD-the SVR forecasting model,and effectively improves the prediction accuracy.Through the forecast figure,we can see that the accuracy of the prediction has been significantly improved,in view of the predict data,and policymakers can effectively take corresponding measures.