This research investigates the impact of retrofit interventions on the energy performance of domestic buildings in Ireland using predictive machine learning (ML) models. The study applies machine learning models to classify the Building Energy Rating (BER) of dwellings in County Dublin, Ireland. With a focus on feature selection in a highly correlated dataset, the study predicts energy ratings with an accuracy of 69%. The Light Gradient Boosting Machine classifier is observed to achieve the best performance among more than twenty ML models applied for prediction. The study also performs retrofit experiments on dwelling features and evaluates their effectiveness in improving energy performanc... More >