Hi KNIME Commuity,
So I have a workflow for Random Forest regression with time series data and contain around 30k Unique ID and each Unique ID have 12 month data. I create a loop to predict the next month data for each Unique ID.
Now I have confusion how to save the model along with sub model for each Unique ID? so I can use it to another workflow
You should be able to write a model using the model writer node:
And read it back in with the Model Reader Node:
You’ll have to work out a naming convention for the file names that you use to save the models - probably best to use your Unique ID.
You can then set up your prediction workflow to load the model as per the Unique ID and to use to to predict.
On a high level, w/o the loop and working out the names, it could look like this:
Here’s the workflow although no data in it…
ModelReaderWriter.knwf (80.4 KB)
