05/15/2023, 3:02 PM
Has anyone tried week numbers (sin & cos) as exogenous variables with ARIMA? Does it generally improve the forecast?
fede (nixtla) (they/them)
05/17/2023, 8:59 PM
! That might be a good approach, but in the end, it depends on the data. here’s an interesting reference on the topic:
Also, you could use the MSTL model to capture the seasonal components of your time series.
05/17/2023, 9:00 PM
@fede (nixtla) (they/them)
much appreciated! Thanks!