Visual Data Mining System Design And Implementation Base On Data Warehouse

In recent years, as information technology widely used in various industries,accumulate more and more data, It is hoped to have the right to the analysis of large amounts of data on the basis of scientific research, decision-making or commercial enterprise management. But what we have today is difficult to data analysis tools for data-depth processing, can hardly be interested in knowledge, Formation of rich data and lack of knowledge of the situation, data mining is to solve the shortcomings of traditional analysis method, and to address large-scale data processing and analysis appear.Currently, academics presented a lot of data mining algorithms, many IT companies have also launched their own data mining products. But these products have problems in poor motivation, low intelligence, systems integration difficulties and the results difficult to understand such defects. This paper gives a visualization data mining based , Analysis of the system architecture and the model of scalability, c, the defects users of the system to reduce dependency. The system consists of data pre-processing, data visualization, modeling and visualization subsystems mining algorithm, is a function of the various components and structures, and discussed the system, the algorithm to achieve key technical problems. The following are the main points : the first chapter on data mining technology background, significance and Research. The second chapter describes the relevant data mining and data warehouse, ETL technologies and data mining process. Chapter III of the design and realization of the applied field of mobile visual data mining system to learn and study the data summary clustering, association rules found that the sequence patterns found that dependence or reliance on models found that the anomalies and trends found, and so on., Discriminant Analysis (Bayesian estimation, and discriminant Fischer, nonparametric test), cluster analysis (system clustering, dynamic clustering), exploratory analysis (PCA analysis, Correlation analysis etc.). Neural Network include: neural network (BP), self-organizing neural networks (self-organizing map, Competitive Learning), and so on. than realize the classification algorithm, clustering algorithm, application templates designed to facilitate future system expansion. Chapter IV presents the data visualization and model visualization, makes no professional background in data mining techniques to the use of data mining for business decision-making possible. Finally, the paper sum up work and made after further research.