H2O production scoring
Models achieve their business value when they are put into production to make predictions (or generate unsupervised results for an unsupervised class of problems). We discuss, in this section, a more detailed view of the H2O pipeline from model building to production scoring.
End-to-end production scoring pipeline with H2O
Take a look at the following diagram showing an end-to-end H2O pipeline from model training to model deployment and production scoring:
Figure 9.3 – High-level view of full scoring pipeline with H2O
Typically, model building is considered a development (DEV) environment, and model scoring is a PROD environment with source data from each respective environment.
For DEV, we have treated feature engineering and model training (and many associated steps such as model explainability and evaluation) extensively in Section 2, Building State-of-the-Art Models on Large Data Volumes Using H2O. We also briefly...
