Choices made in algorithm design such as hyperparameter settings are justified and documented and traced back to the functional and non-functional requirements; the measures of performance used to optimise the design are justified and documented; and evidence shows the design is optimised for performance in the expected operational context rather than against training and test data alone, for example by demonstrating that overfitting was avoided.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.