Research on the Identification Algorithm of Ancient Glass Relics

The burial environment exerts a significant influence on the weathering process of ancient glass, resulting in alterations in composition ratios due to extensive reactions between internal and environmental elements. These modifications have implications for the accurate categorization of ancient glass artifacts. This paper aims to address the problem of analyzing and identifying components of ancient glass using data from Question C of the 2022 National College Students Mathematical Contest in Modeling. By considering factors such as composition, decoration, color, etc., we classify glass relics into categories and examine their chemical component relationships. In this study, OLS regression is utilized to analyze factors that impact cultural relic weathering; clustering algorithms, random forest classification algorithms, global sensitivity analysis techniques, and Chi-square trend tests are applied for composition analysis and identification of ancient glass relics. These methodologies offer novel insights into the identification of ancient cultural relics.