Running time series for multiple brands

Question

Hello Everyone,

I have data for daily sales for different brands-A,B,C,D,E,F,G,H  .  I'm using time series to forecast next 3 months sales.  Every time select each brand and run the process is too lengthy and taking time .

Here are some points I need help on :

1. How shall I run time series modelling for multiple brands at same time ?

2. Also for each brand

*  first I'm running ETS and ARIMA models 
* Then using  TS compare tool selecting model for forecast.
* Which is the best criteria to select models ?
* for Some models my RMSE,ME values are lower for ETS and for some are lower for ARIMA.

3. How can I automate this process or what should I use so that model will directly select best from the ETS and ARIMA and then forecast the values using best model (ETS/ARIMA).

Attached the sample set of data

Please help me with this.

Sample data for time series.xlsx

AngelosPachis · Answer

Hi @ash25sumbre ,

Regarding point no.1, I would definitely recommend reading the following blog post in the community:

https://community.alteryx.com/t5/Alteryx-Designer-Knowledge-Base/Time-Series-Factory-Tools-vs-Batch-Macro-Time-Series/ta-p/455753

It compares batch macros to TS Factory tools, which I believe are the answer to your question.

For point number 2, commonly the models with the lowest RMSE are those yielding the best results.

Hope that helps,

Angelos