Research Of Ancient Ceramic Classiifcation Based On Multivariate Sample Graph

In China, there are many excellent ancient ceramic products which are matchless in the world. And some documents said the strict production process of ancient ceramic products which well consisted of porcelain bodies, glazes and blue-and-white materials with high quality and skillful craft. The unique charm of ancient ceramic products lets them very popular around the globe, so there are full of fakes and imitations everywhere from Qing dynasty, Ming dynasty to the present. Yet the identification of Ancient ceramic fakes is usually done by the method that the archaeologists compare the ceramics under test with the specimens excavated from some ancient kilns and draw conclusions. Due to a long history of Chinese ceramic production and a wide variety of ancient ceramic products with high output, while the essences of learning and meriting from each other in the production process around the ancient kilns from different parts of China, and coupled with war and other historical reasons, resulting in large-scale migration of ceramic artisans who put the firing technology of certain ancient kilns somewhere to another area. Thus, as mentioned before, ancient ceramic products of some different regions are too tough to be identified for their producing areas and authenticity by those usual methods.In recent decades some related agencies have carried out relevant researches on the identification of ancient ceramic products by the adoption of technological approaches, and this main idea is that the manufacturing materials and processing crafts of ancient ceramics of different kinds of ancient kilns and dynasties are not consistent. So the trace element composition and microstructure of each ancient ceramics must have their own information about fired era and location, and then must have the obvious characteristics of be out of artificial control, etc. As a result, these technological approaches can be a useful method to treat the problems of place decision and period decision of ancient ceramic specimens.But in China, the microelement composition of ancient ceramics is very complicated. If any one of the trace element compositions corresponds to a point of hyperspace when each element corresponds to a coordinate axis, so it is very difficult to describe or compare the similarities and variations of these compositions. In response to these problems, and according to the geographic characteristic of the trace element composition’s distribution model, this paper use statistical analysis methods to classify the ancient ceramics specimens in the analysis and determination of their trace element content data. We study the method of multivariable diagram and achieve the practice of trigonometric polynomial graph, constellation diagram, higher chart, tree image with programming by the mathematical software Matlab and R.These practice is applied the classification of ancient ceramic samples, and we make the visualization of these classification results. The findings of this paper are useful to the judgment on the producer and authenticity of ancient ceramic samples.