This document discusses using machine learning algorithms to predict National Achievement Test (NAT) scores of public schools. It analyzes key factors that affect NAT results like internet connection, power distribution, class size, school type, region, budget, and poverty incidence. Different machine learning models are tested on an elementary and secondary school dataset. The best performing models were XGB Classifier and Random Forest Classifier. Key factors found to influence NAT scores were region, poverty incidence, class size, and school type. The document recommends further study of the models, establishing a data culture, and using data science to inform education policies.
Related topics: