Hi there, Just KNIME It! folks! We’re back with a supply chain challenge today.
Claudia, the CEO of a small supermarket chain in the US, is worried about stockouts: a situation that occurs when customer orders exceed the available inventory of an item. She wants to know which warehouses are likely to suffer the most from stockouts, and which types of items are going to be the most problematic. Can you create a model to forecast stockouts for the different warehouses and item types, allowing her to interactively check the forecasts?
Here is the challenge. Let’s use this thread to post our solutions to it, which should be uploaded to your public KNIME Hub spaces with tag JKISeason4-13 .
Need help with tags? To add tag JKISeason4-13 to your workflow, go to the description panel in KNIME Analytics Platform, click the pencil to edit it, and you will see the option for adding tags right there. Let us know if you have any problems!
I used a Random Forest Regression. Many other models could be used (ARIMA, SARIMA, LSTM, …). I wanted to use LSTM, but the extension is not installed yet on my new KNIME instance
Anyway, the concept stays similar.
The solution I provide do the forecast on the fly when the settings are being changed (warehouse or item type change). If it start to be slow or a lot of data needs to be processed, it might be interesting to extract the model training out of the component (and maybe train a generic one to be used, or multiple more specific ones saved somewhere).
The overall workflow:
Inside the component (where the logic is in this one - cf. above if optimization is needed).
And the output as follow, giving some information about model quality and the future prediction of the stockout. If the number of sample is too small, it is being catch with a try-catch logic, including a warning message.
Hi there, Just KNIME It! folks! We’re back with a supply chain challenge today. 
Claudia, the CEO of a small supermarket chain in the US, is worried about stockouts: a situation that occurs when customer orders exceed the available inventory of an item. She wants to know which warehouses are likely to suffer the most from stockouts, and which types of items are going to be the most problematic. Can you create a model to forecast stockouts for the different warehouses and item types, allowing her to interactively check the forecasts? 
Need help with tags? To add tag JKISeason4-13 to your workflow, go to the description panel in KNIME Analytics Platform, click the pencil to edit it, and you will see the option for adding tags right there. 
