Challenge #24: ARIMA Time Series

[myastarling]

This helped take the trepidation out of the ARIMA tool for me! Super fun!

[Greg_Murray]

Solution attached

[ZenonH]

Still clearing out solutions that I started but never finished a long time ago. Playing with ARIMA tools instead of writing out R code is much easier AND more dynamic (:

Saqueador

[klyuka]

Please, more examples like this. Thanks to editor.

[Johnny_Analytics]

The output inside the comment box for the exercise doesn't match the output from the workflow of the solution file.  

Saqueador

I tried a few manually configured ARIMA models to compare against the Auto model.  I found a manually configured (011)(011) model to be a better fit based on MASE (lower is better) vs the Auto generated model.  However, the exercise is looking for the auto-generated output.

[Mike_Hinz]

Had a similar approach to the actual solution.

[bnelly1987]

Not sure why the answer is slightly off from the provided answer, but it seems to be that way in the solution file as well. Mine matches that.

Saqueador

[ponraj]

Here is my solution. 

Saqueador

[JonathanAllenby]

Here's my solution to challenge 24. 
Had to look up the tool required to actually interpret the model produced by the ARIMA tool.

Saqueador

[Marcel_Gavrila]

Here's my solution! I don't know why my numbers are a little bit different...

Saqueador