Question 19
UnclassifiedWhat does a model card primarily document for a machine learning model?
Correct answer: B
Explanation
A model card documents key facts about a machine learning model, including its "intended use," the "training data" it was built on, and its "performance across subgroups." It also records "ethical considerations" and "limitations" so users can assess risks, bias, and appropriate use.
Why each option is right or wrong
A. Server hardware specifications, storage capacity, and deployment rack details.
B. Intended use, training data, performance across subgroups, ethical considerations, and limitations of an ML model.
Model cards are a standard documentation artifact introduced by Mitchell et al. (2019) to summarize a model’s purpose, the data used to train it, and evaluation results, including subgroup performance. They also explicitly note ethical considerations and known limitations so downstream users can judge whether the model is appropriate for a given deployment context and where bias or failure modes may arise.
C. Only the source code and programming language used to build the model.
D. The network topology and routing configuration of the system hosting the model.