The 365 Data Science team is proud to invite you to our own community forum. A very well built system to support your queries, questions and give the chance to show your knowledge and help others in their path of becoming Data Science specialists.
Anybody can ask a question
Anybody can answer
The best answers are voted up and moderated by our team

Reagarding ARIMA model selection

Reagarding ARIMA model selection


How do you select model from [ARIMA(1 1 1),ARIMA(6 1 6)] i.e ARIMA(6 1 3) and ARIMA(5 1 1)?You go through each model from ARIMA(1 1 1) to ARIMA(6 1 6) or their is any other method to select the perfect model? I simply want to know,  if I don’t  want go through each model then what’s the other way to do this?Because it’s very time consuming to go through each model.

1 Answer

365 Team

Hey Ganesh, 
Thanks for reaching out!
Usually finding the model is a trial-and-error approach, where you try different models to determine which fits best. Knowing if the data is stationary (no integration), or experiences no seasonal tendencies (no seasonal orders) reduces our search for the best model. 
Additionally, the ACF and PACF can give us suggestions to the number of lags to include to limit going through overly-simplistic or overly-complex models. 
That being said, later in the course we introduce the auto_arima method which goes through all the models with certain specifications and returns the one which fits best based on a single characteristic (the AIC).  
365 Vik

Complete Data Science Training
Save 60%