Deploying a Gradient Boosted Forest
To deploy the classification model and preprocessing along with it, we’ll use KNIME Analytics Platform’s Integrated Deployment feature. As you will see in the following figure, the preprocessing of the test set along with the Gradient Boosted Trees Predictor node lie directly between the Capture Workflow Start and Capture Workflow End nodes. Any nodes between the capture start and end will be combined into an automatically generated deployment workflow as we will see.
In the following figure, we see how the Integrated Deployment nodes surround the preprocessing and model prediction.
Figure 8.15 – Captured portion of training workflow
Looking at Figure 8.15, you’ll notice that the trained Gradient Boosted Forest model that plugs into the Gradient Boosted Trees Predictor node is also required to execute this section of the training workflow. KNIME’s Integrated Deployment functionality...
