Classification based on Data Envelopment Analysis and supervised learning: A case study on energy performance of residential buildings

Data Envelopment Analysis is an optimization technique for estimating the comparative performance of organizations or decision making units (DMU's). In this study, we use DEA to evaluate the relative energy performance of residential buildings, establishing two classes, efficient and inefficient, based on the DMUs' relative efficiencies as calculated using DEA. We then provide a comparative analysis of the results obtained on classification using the machine learning algorithms: Logistic Regression, Support Vector Machines and Nearest Neighbors validated using K-fold cross validation technique.