AI0-001 Implementing AI Solutions Practice Question
A retail bank deploys a machine learning model that scores loan applications. Compliance requires that the bank be able to explain to regulators why any individual applicant was denied, in terms of the applicant's own feature values. The model is a gradient-boosted tree ensemble trained on 200 features. Which approach BEST satisfies this requirement?
⚠ Common exam trap
The trap here is assuming that any interpretability output, such as a global feature importance chart, counts as an explanation for an individual decision.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Apply SHAP (SHapley Additive exPlanations) values to produce per-applicant feature attributions for each decision.
The requirement is individualized, feature-value-based justification for each denial on a complex ensemble. SHAP attributions decompose a single prediction into additive feature contributions, which is exactly what an adverse-action explanation needs. Global importance, coefficient tables, and raw input logging all describe population behavior or stored data rather than the reason for one applicant's specific outcome, so they cannot satisfy the regulator's question.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Apply SHAP (SHapley Additive exPlanations) values to produce per-applicant feature attributions for each decision.
Why this is correct
SHAP values come from cooperative game theory and assign each feature a signed contribution to the model's output for that specific instance, so a denial can be explained as a sum of feature-level reasons drawn from the applicant's own data. This is model-agnostic, works for tree ensembles, and directly produces the individualized, feature-value-based justification regulators are requesting.
- ✗
Log the model's predicted probability alongside the applicant's raw input record for each decision.
Why it's wrong here
Recording inputs and the output score is necessary for auditability but explains nothing: it shows what was fed in and what came out, without attributing the outcome to specific features. A regulator asking why approval was denied receives no causal or contributive breakdown. This is data retention, not model explanation, and it leaves the compliance gap open.
- ✗
Report the global feature importance ranking produced by the tree ensemble's gain-based split scores.
Why it's wrong here
Gain-based global importance summarizes which features matter across the whole training population, not why one particular applicant was denied. It cannot tell a regulator that this applicant's debt-to-income ratio contributed so many points against approval. Global rankings also hide sign and magnitude of effects, so they fail the individualized explanation requirement entirely.
- ✗
Retrain the model as a logistic regression and present the learned coefficients as the explanation.
Why it's wrong here
Coefficients are global and constant across applicants; they describe average log-odds effects, not an individual's decision path. Retraining also discards the ensemble's accuracy and forces a full revalidation cycle. While intrinsically interpretable models are useful, this scenario asks for an explanation of the existing deployed model's decisions, and coefficient tables do not provide per-applicant reasons.
About these practice questions
One of 962 original AI0-001 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official CompTIA exam blueprint
This AI0-001 practice question is part of Courseiva's free CompTIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI0-001 exam.