Contextually-enhanced transformers integrating timetable and occupancy data achieve 26.6% and 56.3% average MAE reductions in railway and building energy forecasting respectively, outperforming prior methods.
Smart-Meter Big Data for Load Forecasting: An Alternative Approach to Clustering
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CY 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Integrating the Expected Future in Load Forecasts with Contextually Enhanced Transformer Models
Contextually-enhanced transformers integrating timetable and occupancy data achieve 26.6% and 56.3% average MAE reductions in railway and building energy forecasting respectively, outperforming prior methods.