A study on the chemical composition of ancient glass products based on gray correlation analysis

Abstract

The ancient high potassium and lead-barium glass made locally at the end of the Warring States period in China not only represents the trade interactions along the Silk Road in China, but also is the crystallization of science and art of the ancient people, and has high research value. In this paper, we first adopt the systematic hierarchical clustering based on Euclidean distance, by dividing different degrees of Euclidean distance to obtain the statistical law of glass type classification. Based on this, a hierarchical clustering model based on PCA dimensionality reduction is used to classify the chemical composition into subclasses. Then, the unknown glass-types were identified according to the classification law, and a classification model based on decision tree cross-validation was established. To test the reliability of the model, we used a comparative analysis with the mean line graph of each component before and after known weathering, and the results were finally found to be reliable. Finally, we used gray correlation analysis to discuss the classification of the two types of glass artifacts, solved the correlation degree of fourteen elements in the two types of glass separately, and analyzed the differences between them.