基于多元回归和聚类分析的古代玻璃制品成分分析与鉴别

Composition Analysis and Identification of Ancient Glass Products Based on Multiple Regression and Cluster Analysis

摘要: 本文主要研究古代玻璃制品在长期埋藏环境影响下不同类型玻璃文物表面风化前后化学成分比例之间的关系并进行相应的预测与鉴别。通过建立相关性分析和偏最小二乘回归模型进行各化学成分间关系分析，结合持时放大比例进行风化前化学成分比例预测；同时建立聚类分析模型进行文物分类并鉴别。此研究在现实物理化学意义的角度分析问题，有助于古代玻璃制品出土后的成分分析与鉴别。

Abstract: This paper mainly studies the relationship between the chemical composition ratio of different types of glass cultural relics before and after surface weathering of ancient glass products under the influence of long-term burial environment, and makes corresponding prediction and identification. Through the establishment of correlation analysis and partial least squares regression model to an-alyze the relationship between various chemical components, combined with the time-sustained amplification ratio to predict the proportion of chemical components before weathering; at the same time, a cluster analysis model was established to classify and identify cultural relics. This study analyzes the problem from the perspective of realistic physical and chemical significance, which is helpful for the composition analysis and identification of ancient glass products after unearthed.