An Empirical Study on ML Models with Glass Classification Dataset

Glass classification in machine learning is an important task with various practical applications, including quality control, material identification and forensic analysis. The chosen glass classification dataset contains information about the properties of different glass samples and the goal is to classify these samples into different types of glass based on their features. This paper presents a summary of how different machine learning algorithms (ML) perform to the particular conditions presented by us. Different ML algorithms were modeled including Logistic Regression (LR) , Support Vector Machines (SVM), Decision Tree, K-Nearest Neighbors (KNN), XGBoost, Naive Bayes, Random Forest and also the Multi-Layer Perceptron. Finally we found that tree boosting was a highly effective and widely used ML method and with that, lazy - supervised learning has come to the lime light due to its ability to work very differently yet effective. On the basis of this study, we also describe the similarities between K-Nearest Neighbors and XGBoost classification algorithms.