An Adaptive Transfer-learning Based Missing Data Imputation For Reliable Accounting Of Building Carbon Footprint.

Abstract

Buildings generate nearly 40% of annual global GHG emissions. In recent years, many organizations set goals to develop Zero-energy buildings by the next decade. But most of the organization is facing a challenge in getting high-quality building level energy consumption data. To address this challenge, we developed an adaptive transfer learning-based methodology to impute the different rate of missing energy consumption data with over 90% accuracy.