Identification of Glass Relics Based on Cluster Analysis and Grey Correlation Degree

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Ancient glass artifacts often undergo weathering in burial environments, leading to element exchange and composition changes that affect their identification and classification. This study employs multiple linear regression, grey relational analysis, and cluster analysis to examine glass weathering and classification. Statistical analysis shows high‑potassium glass has stronger weathering resistance than lead‑barium glass, while color has little influence. Multiple linear regression estimates the original chemical composition before weathering. Cluster analysis subdivides high‑potassium glass and lead‑barium glass into distinct subgroups based on key oxides, with good model sensitivity verifying the classification. Grey relational analysis further reveals compositional relationship differences between glass types. This research provides reliable scientific support for the compositional identification, classification, and conservation of ancient glass artifacts.

This publication has 8 references indexed in Scilit: