Darius Marginean12/22/2022, 5:22 PM
do we need to do some scaling on data?
fcst = MLForecast( models=models, freq='W-MON', differences=data['differences'] )
models = [lgb.LGBMRegressor()] fcst = MLForecast( models=models, freq='W-MON', differences= )
Max (Nixtla)12/22/2022, 5:39 PM
José Morales12/22/2022, 6:13 PM
Darius Marginean12/22/2022, 6:14 PM
José Morales12/22/2022, 6:18 PM
Darius Marginean12/23/2022, 10:13 AM
the output of the predictions wasn't the same on the whole horizon. We've set
differences = 
as a suggestion from @fede (nixtla) (they/them) in a previous thread to capture the yearly trend & seasonalities.
differenences = 
José Morales12/28/2022, 2:35 AM