Hi there,

I have a question related to the combination of multi-timeseries forecasting and multivariate series forecasting.
I have data of historical sales and discounts for several products. I want to predict future sales for multiple products (= independent multi-timeseries) using multiple timeseries as input: past sales and past discounts, they highly influence each other (= dependent multivariate series). I would prefer not to build a multivariate model per product as I expect products to learn from each other, that is why I want to use one global model.
From your documentation I cannot seem to find an example where both can be combined. Is there a way to do this?

Thanks in advance!