Research on Classification of Ancient Glass Products Based on Machine Learning

In the process of weathering for a long time, due to the exchange of internal elements and environmental elements, the composition proportion of glass cultural relics changes, which affects the correct judgment of archeological workers on their categories. In this paper, by machine learning method, the random forest algorithm is used to find out the feature attributes with higher importance. Based on the results obtained by random forest, the decision tree algorithm is used to get the decision tree and the classification method of high-potassium glass and lead-barium glass. In addition, the clustering coefficients of high-potassium glass and lead-barium glass were calculated by clustering algorithm, and the number of subclasses of the two kinds of glass was calculated. Then the random forest and decision tree algorithms are used to get the classification methods of high-potassium glass subclass and lead-barium glass subclass.