Composition analysis and identification of ancient glass objects based on Spearman correlation analysis model

Abstract

With the rising fervor of Belt and Road, it has drawn attention and research to the ancient Silk Road commodity trade exchanges, of which ancient glass products are valuable physical evidence of the early commodity trade exchanges. Now the chemical composition of ancient glass is analyzed and identified, which helps to find out the key and know-how of ancient glass products refining technology. In this paper, we establish a chi-square test and Spearman correlation analysis model to quantitatively analyze the relationship between weathering and type, decoration and color to establish a weathering interval estimation model, so as to predict the chemical composition content before weathering. Logistic dichotomous classification model, BP neural network model, BP neural network model optimized by genetic algorithm and random forest classification prediction model were established respectively to predict three high potassium glasses and five lead-barium glasses out of eight compounds.