Study on Chemical Composition Analysis and Methods of Ancient Glass Relics

Abstract

In this paper, the relationship between the surface weathering of glass artifacts and the type, decoration, and color of glass was visualized by using a columnar table analysis[1], and the relationship between the three factors on the surface weathering was tested separately for significance by using a chi-square test; in addition, a multi-factor ANOVA was used to analyze whether the three factors of the type, decoration, and color of glass and their interaction on the surface weathering of glass artifacts had The significant relationships were analyzed by multivariate ANOVA. Then, the relative contributions of different types of glass to the prediction accuracy of the model were ranked by importance according to the proportion of chemical components before and after weathering at each sample point through random forest, and the chemical components were selected according to the importance ranking, and after removing most of the unimportant chemical components, the final retained chemical components were used to construct the final random forest regression model to obtain the laws of their classification, and finally the corresponding The results were obtained. The chemical components such as calcium oxide, copper oxide, iron oxide, and nitrogen oxide were drawn for visual analysis according to their importance, and we compared the derived results with the known literature and found that they were consistent with the results classified by the known literature.