| The securities market is an important part of the market economy, development of the securities market is not only related to the country's economic prosperity, but also the income of the individual investor. Therefore, forecasting the stock trend effectively have great theoretical significance and practical value. However, the impact of various factors by the community, the quotations of the securities market and the rules are very complex, using traditional techniques to forecast potential stock, the result is not satisfactory. The main study of this article is to use data mining methods to tap potential stock in securities market.With regard to application of data mining technologies in the security analysis, domestic and foreign experts and scholars have done a lot of research, but still at the exploratory stage, lack of instructive research for practical application. In view of having the characteristics of nonlinear, multilevel and others for the stock information, the article summarized the results of our predecessors, system research on application of data mining of potential stock based on BP neural network of data mining technology, verifying accuracy of the forecast through analysis of examples, consequently demonstrate feasibility of the application on the stock forecast. In addition, for the stock information also has ambiguous, incomplete, continuous and hidden characteristics, this article introduce the grey system theory to do further study, through complementary analysis of the grey system theory and BP neural network, combined of both, proposed a new forecast method of grey BP neural network, accomplish improvement of BP neural network. This article combine GM model with BP neural network to build a new forecasting model and applied on forecasting of stock price, verified by examples on accuracy of forecast have greatly increased by improved grey BP neural network, consequently show its progress in the mining of the potential stock. |
PDF Full Text Request