Work

Predicting Power Dispatch for Unit Commitment Problems Using Graph-Temporal Convolutional Networks With Constrained Learning