Research on The Classification Models of Ancient Chinese Glass Based on K-means Clustering Algorithm

Abstract

Ancient Chinese glass is mainly potassium glass and lead-barium glass. Based on the main chemical composition of potassium glass and lead-barium glass, a K-means clustering algorithm was used to establish classification coefficients to classify glass types at one level, with the classification condition being whether the classification coefficient ε was greater than 5/16, with greater than being potassium glass and less than being lead-barium glass. a clustering algorithm with a K value of 2 was then used to classify the same type of glass into subclasses, with potassium glass being classified as whether it contained Na2O and lead-barium glass being classified as the BaO content is greater than 20% and smaller.