Kejela, Girma (Master thesis, 2012)
Although state-of-the-art models like linear regression and neural networks have been widely used for electricity consumption forecasting, the demand for improved prediction accuracy is still very high. A small improvement ...