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.
Ask
Anybody can ask a question
Answer
Anybody can answer
Vote
The best answers are voted up and moderated by our team

How Computationally expensive is Auto Arima Model

How Computationally expensive is Auto Arima Model

0
Votes
1
Answer

I am currently learning about Auto Arima models, but once I ran my model based on the template you provided, it takes forever to compile, I just copied the relevant code to Google Collab and although it took a while, it was much faster, I use a HP Spectre,(No GPU, Intel Ice Lake i7) and on Google Collab I used a GPU accelerator, is using Google Collab more advisable now?

And generally speaking what are the recommended specs to get results from Auto Arima in decent amount of time

1 Answer

365 Team
0
Votes

Hey Aditya, 
Thanks for reaching out!
Yeah, auto_arima is very computationally expensive because it runs many different models 1 by 1 based on the arguments given. Hence, the more we know about the data, the more we can limit the arguments we pass to the method. By doing so, we can reduce the number of models auto_arima fits and decrease running times. 
 
And generally speaking what are the recommended specs to get results from Auto Arima in decent amount of time.
Honestly, your specs seem fine for the task, but it all depends on what you mean by decent. The Auto ARIMA can sometimes even take 20 minutes for models with numerous exogenous variables, so maybe switching to Google Colabs is the right move. 
 
Best, 
365 Vik