Template Credit: Adapted from a template made available by Dr. Jason Brownlee of Machine Learning Mastery.

SUMMARY: The purpose of this project is to construct a predictive model using various machine learning algorithms and to document the end-to-end steps using a template. The Glass Identification dataset is a multi-class classification situation where we are trying to predict one of several (more than two) possible outcomes.

INTRODUCTION: The dataset involves predicting, from USA Forensic Science Service, six types of glass; defined in terms of their oxide content (i.e. Na, Fe, K, etc). The study of classification of types of glass was also partly motivated by criminological investigation. At the scene of the crime, the glass left can be used as evidence…if it is correctly identified!

ANALYSIS: The baseline performance of the model achieved an average accuracy score of 65.00%. After tuning the hyperparameters, the best model processed the training dataset with an accuracy of 70.00%. Furthermore, the final model processed the test dataset with an accuracy of 68.52%, which indicated that we might have a variance problem. We need to gather more data or apply regularization techniques in training to narrow the variance gap before deploying the model in production.

CONCLUSION: For this dataset, the model built using Keras and TensorFlow achieved a satisfactory result and should be considered for future modeling activities.

Dataset Used: Glass Identification Data Set

Dataset ML Model: Multi-Class classification with numerical attributes

Dataset Reference: https://archive.ics.uci.edu/ml/datasets/glass+identification

One source of potential performance benchmarks: https://www.kaggle.com/uciml/glass

The HTML formatted report can be found here on GitHub.