This message was deleted.
# general
s
This message was deleted.
t
Although I wouldn't implement the global models suggested this paper says that a global model can be better even among unrelated series: https://arxiv.org/abs/2008.00444 Specifically: "Global methods are not more restrictive than local methods, both can produce the same forecasts without any assumptions about similarity of the series. Global models can succeed in a wider range of problems than previously thought." I would add that it depends on your data, if you have like 20 series it is probably better to hand-hold local models with unrelated series. Also, if you have any exogenous features, maybe like temperature or something, that tips the scales towards global models IMO.
s
Thanks