Classification and identification of glass artifacts based on high-dimensional clustering and XGBoost algorithm

Abstract

After the weathering of glass artifacts, a large number of environmental elements are exchanged with elements inside the glass artifacts. In this paper, based on the chemical composition of the artifact samples and other detection means, they are classified into two types: lead-barium glass and high-potassium glass. A sub-classification model based on high-dimensional clustering is introduced, and an identification model is established by optimizing the K-means algorithm. Finally, a feature classification model based on XGBoost algorithm is used to analyze the chemical composition of the unknown class of glass artifacts.