古玻璃成分的分类预测研究

Research on Classification Prediction of Ancient Glass Omposition

基于化学成分的分析,研究出古玻璃中高钾类型和铅钡类型的主导化学成分;根据主成分分析的结果,再进行AGG层次聚类分析将古玻璃分成若干亚类,得到古玻璃分类规律以及亚类划分规律中的判定化学成分及含量指标;分析了风化前后,主要化学成分所呈现的各自不同的变化规律,并对未知古玻璃品类及其风化情况进行了预测;最后基于决策树的预测算法对化学成分分析的结果进行验证。这些分析结果对于保护和进一步研究古玻璃有着重要意义。

Based on the analysis of chemical composition,the dominant chemical components of high potassium type and lead barium type in ancient glass were studied,and according to the results of principal component analysis,AGG hierarchical cluster analysis was carried out to divide ancient glass into several subcategories,and the classification law of ancient glass and the chemical composition and content index in the subcategory classification law were obtained,the different change laws of the main chemical components before and after weathering were analyzed,and the unknown ancient glass species and weathering were predicted.Finally,the prediction algorithm based on the decision tree verifies the results of the chemical composition analysis.These analytical results have important implications for the preservation and further study of ancient glass.