For Merit standard learners will test and review their AI solution and related data sets.
As part of the review learners will consider the quality, accuracy and reliability of the
outcomes generated by the solution, and use these observations to refine the data sets,
and AI approaches to improve the solution.
As part of this process learners should review their objectives and hypotheses and
update these as required based on the outcomes of each iteration of the AI solution.
Learners will produce additional iterations of the solution and data sets, as they review
and refine, which should be included in their portfolio of evidence. The evidence should
be supported by accompanying documentation detailing the review and testing process
they followed.
For Pass standard, learners will produce a set of objectives for the AI project they
intend to implement. The objectives should provide a summary of the overall purpose
of the project and a list of success criteria/functional requirements against which they
review their AI solution and related data sets. The functional requirements will consist
of metrics by which the solution can be measured, and a set of hypotheses that will be
tested/explored.
Learners will select a suitable AI model and use this to produce an AI solution to meet
the objectives they have devised. The solution will be functional but there may be some
minor issues in terms of accuracy and/or performance. The solution will meet some of
the listed objectives that they have listed.
Learners will provide a portfolio of evidence that demonstrates the gathering and
preparation of data, and documents the implementation of the solution. They should
also provide evidence of a working solution and how the AI solution meets the
objectives. How this is presented will vary depending on the AI project chosen.