Analysis and Identification of the Composition of Ancient Glass Objects Based on Logistic Regression Analysis

Abstract

Ancient glass objects were weathered by environmental influences, and the analysis of their main components and identification of their categories is a prerequisite for all subsequent research work. This paper uses Spearman's correlation coefficient and chi-square test of variance to analyses the correlation between weathering and type, color and decoration on the surface of cultural relics; secondly, descriptive statistical tests are conducted separately for two different types of glass objects to obtain the relationship between the changes in the content of each chemical component of different types of glass objects before and after weathering; finally, by analyzing the above statistical laws, a weighted average prediction model is established to The paper analyzed the classification patterns of two types of glass, namely high potassium and lead-barium, by establishing a logistic regression model; the Spearman coefficient and analysis of variance were used to investigate the correlation between surface weathering and each component. A principal component analysis of 14 different chemical components was carried out to achieve the effect of dimensionality reduction and to interpret the meaning of each principal component; a logistic regression model was established to regress the data after the principal component analysis, which showed that the classification criteria were more reasonable and the sensitivity of the model was high.