Analysis and identification of glass components based on independent sample t-test and systematic clustering

Abstract

With the development of archaeology, research on glass weathering has also received much attention. The glass composition was found to be an important factor influencing weathering through a large number of experiments. In this paper, glass composition and glass weathering are analyzed and studied by statistical analysis and integrated algorithm processing of data. The relationship between surface weathering and glass type, decoration, and color was analyzed and processed. Independent sample t-tests using SPSS were used to obtain the distribution patterns of weathering and chemical composition. The mean ratio analysis and BP neural network were then used to predict the pretreated data, respectively. The two results were obtained and statistically analyzed respectively. The classification patterns of high potassium and lead-barium glass were analyzed by statistical analysis and integrated algorithm Bagging model. The chemical composition data of high potassium and lead-barium glass were classified into subclasses and a systematic clustering model was used.