I have created a Linear Regression model, trained it and tested it. I’m happy with the results. I need to use it on fresh new data, this data is in a csv file, and as expected does not include the target variable (the value I’m trying to predict)
I would save the model (Model Writer node) and would build a second workflow, which is often called deployment workflow.
This should first read your csv file and the saved model (Model Reader node). Then it should do the same data preparation as in your training workflow (e.g. strings to number…) before it applies the model with the predictor node.
You can find an example on the EXAMPES Server under:
50 _Applications/38_Credit_Risk_Assessment