Research on Clustering of Glass Cultural Relics Based on Weathering Data

This paper mainly studies the classification of ancient glass products through machine learning technology. On the basis of weathering data, this paper studies the classification of high potassium and lead barium glasses according to their chemical compositions, and explores the classification rules according to the classification characteristics of the classification results. Firstly, the K-means++ clustering method is used to explore the classification rules of high-potassium and lead-barium glasses，the results show that the content of weathering chemical components is the main classification standard. Then, random forest is used to screen the key chemical composition combinations in each class, and fuzzy C-means sub-classification is used as an index for each class, the results show that: high potassium type class I is divided into four classes, class II is divided into three classes, lead-barium type class I is divided into four classes, class II is divided into three classes, and class III is divided into four classes. Finally, the above methods are analyzed as a case, which shows that the classification method in this paper has a good performance in the stability and sensitivity of glass cultural relics classification, and the research method in this paper provides a certain theoretical basis and technical support for the protection of ancient glass cultural relics.