These points:
- you might have to reset the H2O workflow so it would use your own paths
- the workflow is based on a sample file of house prices and the validation is based on a few minutes of model time just to illustrate the results. It would do little to solve a real-world problem (that would take much more time and preparation of your data)
- the file is in the subfolder /model/validate of the workflow
- the workflow does use the open-source version of H2O.ai AutoML as described in the article below, there you will also find details about the validation which the mentioned file is a part (about variable importance)
- included in the workflow is a PDF in the /script/ subfolder with documentation “H2O.ai AutoML in KNIME for regression problems.pdf”
- if you want to use the Driverless AI you should follow their installation instructions and connect it to KNIME (check out the links provided or the examples 1|2)
In general it would make sens to familiarize yourself with the concept of KNIME file handling and paths