The research was conducted by me and Menna Nawar using the ASHRAE dataset provided by kaggle.
You can find the Paper here on arxiv.
- introduced a new transfer learning method that can help with forecasting the energy consumption of buildings with scarce data
- compared our method with the well-known machine learning and deep learning models used for timeseries forecasting
- compared the performance of the before/after using our introduced technique