Glass classification and identification based on lasso regression and K-means clustering

Abstract

Ancient glass is classified into two types, high potassium glass and lead-barium glass, which are highly susceptible to weathering by the burial environment. In order to protect the glass artifacts more safely, based on some data related to the chemical composition ratio of glass artifacts,by the lasso regression model based on the logit transformation, the classification rules of high potassium and lead-barium glass were analyzed, and the glass was divided into four subclasses based on the principal component analysis using K-means clustering, and sensitivity analysis was performed on the classification results. Finally, the validity of the model was verified by determining the type of glass to which a set of test data belonged.