Hi all,
I am completely new to time series models and hope you can help me with this, since I have no idea where I need to start.
I would like to predict the daily sales per product. Therefore I have a dataset that contains the sales of 730 days for each product. I also have additional variables such as price and if the product was in a promotion for example. The dataset looks like this:
Date Product Sales Price Promotion
2018/01/01 Product 1 10 10,00 No
2018/01/01 Product 2 12 14,00 No
2018/01/01 Product 3 40 18,00 Yes
2018/01/01 Product 4 30 12,00 Yes
2018/01/01 Product 5 22 8,00 Yes
2018/01/02 Product 1 10 10,50 No
2018/01/02 Product 2 40 12,00 Yes
2018/01/02 Product 3 41 18,00 Yes
2018/01/02 Product 4 14 11,75 No
2018/01/02 Product 5 20 8,50 Yes
For my prediction I want to compare three types of models: exponential smoothing (as the baseline), multiple linear regression (as a more simple model) and SVR (as a more complex model).
I read a lot about time series modelling, but really don't know where to start when predicting the sales with more than one product, since I do not want to manually do a time series forecast for each product. I saw something like the HTS package for hierarchical time series, but saw that this only supports ARIMA, random walk and exponential smoothing methods. I was also thinking that maybe a dummy column for my products would suffice? I am currently a bit stuck at the moment so I'm hoping someone can tell me what kind of packages I need to use, maybe give me some example of codes etc., or can tell me what kind of search terms I can use on Google so that I can find relevant information for this problem.
Thanks a lot in advance!
Kind regards,
Odette
Referred here by Forecasting: Principles and Practice, by Rob J Hyndman and George Athanasopoulos