Research on Glass Classification and Identification based on Machine Learning and Monte Carlo Algorithm

Abstract

In order to classify and identify glass, we used logistic regression and various machine learning algorithms to classify glass of known composition and class, and used various metrics to evaluate the classification results comprehensively. To further improve the reliability and accuracy of the classification. In order to further improve the reliability and accuracy of the classification, we used grid search and Bayesian optimization to adjust the parameters of the SVM, and the optimized SVM was able to classify the glass with known components and classes with 99.25% accuracy. Finally, we applied the SVMs with grid search and Bayesian optimization to predict the type of glass with only the chemical composition identified and the category unknown.