A study on the composition analysis and identification of ancient glass products

Abstract

The study of ancient glass, as an important branch of scientific and technological archaeology, is an important physical material for exploring the economic, technological and cultural exchanges between China and foreign countries on the Silk Road. It is very important for studying the development of ancient society and cultural exchanges between China and foreign countries. In this paper, we focus on the composition analysis and identification of ancient glass products, based on data mining in the annexes and with the help of software analysis, we give the weathering pattern of glass surface, predict the chemical composition content before weathering, as well as cluster analysis and correlation analysis of different types of glass and their chemical compositions. In this paper, a sample of glass types and the correlation of their chemical compositions is established for analysis. Based on this, a decision tree classification model was built to classify the artifacts in the sample and determine the magnitude of the chemical composition of the classification criteria. The Elbow method (elbow rule) is used to determine the optimal number of aggregated classes k, and K-Means clustering analysis is performed, and then the decision tree classification model is used to quasi-determine and classify the subclass classification criteria for the clustered results. Finally, the XGBoost model is trained to achieve the prediction of the given sample types.