The days with the largest error, with hints generated automatically from the prices themselves. They are clues, not a root-cause analysis: the model does not see wind, sun, gas or plant outages.
Shown is the model with the lowest D+1 error over the last 30 days.
lear.v1, D+1: MAE 55 EUR/MWh
lear.v1, D+1: MAE 51 EUR/MWh
lear.v1, D+1: MAE 39 EUR/MWh
lear.v1, D+1: MAE 38 EUR/MWh
lear.v1, D+1: MAE 37 EUR/MWh
lear.v1, D+1: MAE 37 EUR/MWh
lear.v1, D+1: MAE 31 EUR/MWh
lear.v1, D+1: MAE 30 EUR/MWh
lear.v1, D+1: MAE 93 EUR/MWh
lear.v1, D+1: MAE 83 EUR/MWh
lear.v1, D+1: MAE 74 EUR/MWh
lear.v1, D+1: MAE 73 EUR/MWh
gbm.v1, D+1: MAE 79 EUR/MWh
gbm.v1, D+1: MAE 77 EUR/MWh
gbm.v1, D+1: MAE 74 EUR/MWh
gbm.v1, D+1: MAE 71 EUR/MWh
naive_similar_day.v1, D+1: MAE 160 EUR/MWh
naive_similar_day.v1, D+1: MAE 144 EUR/MWh
naive_similar_day.v1, D+1: MAE 130 EUR/MWh
naive_similar_day.v1, D+1: MAE 124 EUR/MWh