EnergyMamba: An Uncertainty-Aware Graph-Enhanced Selective State Space Model for Energy Consumption Prediction
Researchers have developed EnergyMamba, a novel framework designed to improve energy consumption prediction by integrating spatial dependencies with temporal dynamics. This model utilizes a Graph-Enhanced Selective State Space Model to incorporate grid topology and an Adaptive Sequential Conformalized Quantile Regression module for uncertainty estimation. Evaluations on real-world datasets demonstrate EnergyMamba's superior accuracy and reliability compared to existing methods. AI
IMPACT Introduces a novel spatiotemporal modeling approach for energy prediction, enhancing accuracy and uncertainty quantification.