The Improvement Algorithm And Application Of The Grey Neural Network

With the rapid development of computer technology and database technology, the research on data mining technology more and more in depth and application become increasingly widespread. As a single data mining algorithm, the functionality is limited and there be some shortcomings. Therefore, researchers combine the single algorithms to a new algorithm. The new model absorbs the advantages of single models and overcome the shortcomings, so the combined data mining algorithm model make a better predict performance. Compared to the traditional data mining model, gray neural model is combined by two traditional models that gray model and neural model. On the basis, according to the characteristerics of the single models, we can do the improved and optimized work. Currently, the gray neural network has been used in pattern recognition, data mining and machine learning and other fields widely.This paper describes a variety of traditional data mining algorithms and their key technologies, and several common principles of data mining algorithms are analyzed. Analysis of the advantages and disadvantages of several data mining algorithms is also mentioned. The paper introduces the gray neural network algorithm particularly. Focus on research and analysis of the steps that the algorithm takes and the problems in the prediction process. On this basis, the paper proposes the improvement and solution, established the grey neural network model of GNNM(0, N). Finally, the past years'soil erosion data of Yunxi area of Danjiangkou reservoir for the experimental data, Microsoft Visual Studio C# and SQL Server as the developmental tool. By using the improved algorithm to complete the quantitative prediction of soil erosion, 20% of the improved model training time is saved and the results of improved model is 70% better than the original model. The prediction performance gets much better.