Figure 9
From: Frugal day-ahead forecasting of multiple local electricity loads by aggregating adaptive models
Evolution of (a) state coefficients of a Kalman TL model \((\hat{\theta }_{t,d} - \hat{\theta }_{0,d})_{d=1}^D\) and (b) hybrid state coefficients of a AGG Kalman TL aggregation \((\widetilde{\theta }_{t,d} - \widetilde{\theta }_{0,d})_{d=1}^D\), during the first lockdown and for the same target forecasting task.
