Uncovering Inconsistencies in Building Energy Ratings through Self-Supervised Contrastive Learning

Inconsistent and potentially corrupted building energy assessment data limits the effectiveness of data-driven approaches for predicting fine-grained building energy ratings.