A regulator needs to understand each decision. What should guide your modeling?
Why A is the answer
Interpretability requirements
When a regulator needs to understand each decision, interpretability must guide the modeling choice.
In detail
- Interpretability requirements should guide the choice.
- Regulation can narrow your choices.
Use case
An illustrative scenario, not a real incident.
A lender must explain each refusal, so it chooses a model it can explain.
See it in different ways
mindmap
root((Decay))
Data drift
Inputs changed
Concept drift
Relationship changed
Skew
Training differs from serving
Attribution drift
Feature importance movedflowchart TD
Q{"Must each decision be explained?"}
Q -->|"Yes"| A0["Interpretable model"]
Q -->|"No"| A1["Wider choice"]Key point: Interpretability shapes model choice.
Why the other options are wrong
- B. The largest model that your budget allows
- Large models are harder to explain.
- C. The newest available architecture
- New does not mean explainable.
- D. The shortest training time
- Time is not the requirement.