Research Of Visual Classification Method Based On Multivariate Chart Presentation

In modern society, the magnitudes of data and information are increasing rapidly as the development of science and technology. In order to obtained useful data from large numbers of information, it is very important to be analyzed and taken before processed further. On base of pioneer's achievements, the method of data visualization in term of principle of multi-graph is researched in this paper.First, the concept and principle of multi-dimensional data visualization are introduced theoretically and the basic techniques of it are analyzed. The uses of multi-graph principle in data sort are summarized and the comparisons of the sort method on basis of traditional technique and multivariate chart representation are taken.Second, the assoeted methods of multi- dimensional data visualization coupled with the theory of it and multivariate chart representation are discussed thoroughly. According to the principle of multivariate chart representation, a new sort way that could help the users to understand the data and the assoeted truth is designed.Third, according to the theory of multivariate chart representation, the classification of origin gas, the appraisement of cities and the fault diagnose of bearing are researched by help of radar graph and face chart. The validity of multivariate chart in visualization sort is verified and it may be as reference to multivariate chart representation by pre-processing, reducing the dimension and classifying the data.Finally, the comparison of the results on basis of multivariate chart representation and traditional classification are made and it show that the results obtained from the method of multivariate chart representation is more easy to be understood and accepted.. What is more, the classified method that is proposedin this paper is more applicable and flexible as the applications in the different field are analyzed.