[Models] Add possibility to ensemble different models

[AzulGarza]

Description

Often, some pipelines require the ensembling of different statistical models. We can accomplish this by introducing a new module in the library named statsforecast/ensembles.py, which contains various methods for ensembling models. The general implementation might look something like this:

from statsforecast.ensembles import FFORMA, MedianEnsemble

sf = StatsForecast(
     models=[FFORMA(AutoARIMA(), AutoETS()), MedianEnsemble(Naive(), SeasonalNaive())]
)

Ideally, the output dataframe should include the forecasts from individual models as well as the ensemble method's forecast.

Use case

No response

[hberande]

Hi @AzulGarza, have we implemented above ensemble model in NIXTLA ?

[anthonygiorgio97]

Hi @AzulGarza . It seems to be a really interesting solution. Are there any plans to implement it in the short term?