Skip to content

FRM Part II · FRM Exam Part II · Advances in Artificial Intelligence: Implications for Capital Markets Activities

A risk team must explain to a regulator why a complex ensemble model rejected a loan applicant. The model is accurate but not inherently interpretable. Which approach best addresses the explainability requirement while retaining the model?

The best approach is to apply post-hoc explanation techniques, such as Shapley-value feature attribution, which show how each input contributed to the individual decision. This keeps the accurate model while giving the regulator an explanation. Accuracy alone does not satisfy explainability, and adding trees does not improve interpretability.

  1. ARemove validation since accuracy is high
  2. BApply post-hoc explanation techniques such as feature-attribution (e.g., Shapley-value based) methods to show each input's contributionCorrect
  3. CReplace the target variable with a random one to test stability
  4. DIncrease the number of trees to reduce variance in the explanation

Explanation

Post-hoc attribution methods such as Shapley-value approaches quantify how each feature contributed to an individual prediction, supporting explanations for opaque models. High accuracy does not replace explainability or validation. Randomizing the target is meaningless here, and adding trees does not make the model more interpretable.

Did you get it right without looking?

One question tells you little. A timed set on Advances in Artificial Intelligence: Implications for Capital Markets Activities shows your real accuracy, how long you take and where you lose marks.

Practise Advances in Artificial Intelligence: Implications for Capital Markets ActivitiesFree diagnostic mock

More Advances in Artificial Intelligence: Implications for Capital Markets Activities questions