AIF-C01 Guidelines for Responsible AI Practice Question
Which TWO techniques provide interpretability for machine learning models at a local (per-prediction) level? (Choose two.)
⚠ Common exam trap
The AWS AI Practitioner exam often tests the distinction between global and local interpretability. The trap here is that candidates confuse global techniques like partial dependence plots or permutation feature importance with local methods, because all provide 'feature importance' but at different scopes. Remember that SHAP and LIME explain individual predictions, while PDP and permutation importance explain overall model behavior.
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
✓
SHAP values
SHAP (SHapley Additive exPlanations) values (option A) are correct because they decompose an individual prediction into additive per-feature contributions based on Shapley values from cooperative game theory, giving a local, per-instance explanation of how each feature pushed the model output. LIME (Local Interpretable Model-agnostic Explanations) (option D) is also correct because it fits a simple interpretable surrogate model (e.g., sparse linear model) around a single prediction by perturbing the instance's neighborhood, thereby explaining that specific prediction locally. Partial dependence plots (option B) are not local — they show the average marginal effect of a feature across the entire dataset, a global technique. A confusion matrix (option C) is a global performance summary of classification outcomes (TP/FP/TN/FN counts), not a per-prediction explanation. Permutation feature importance (option E) is a global method that measures the drop in overall model performance when a feature's values are shuffled across the dataset, so it does not explain individual predictions.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
SHAP values
Why this is correct
SHAP values satisfy the per-prediction constraint by computing Shapley values from cooperative game theory, attributing each feature's marginal contribution to a single prediction. Unlike global methods such as permutation importance, which aggregate across the dataset, SHAP explains why one specific output occurred, giving local interpretability.
- ✗
Partial dependence plots
Why it's wrong here
Partial dependence plots average the model's output over the dataset while varying one feature, producing a global marginal effect curve rather than an explanation of an individual prediction. They are tempting because they visualise feature influence, and would be correct when describing overall model behaviour across the population.
- ✗
Confusion matrix
Why it's wrong here
A confusion matrix aggregates classification outcomes across the whole test set, so it describes global model performance rather than attributing any single prediction to its input features. It is tempting because it is a standard evaluation tool, and it would be the right choice when reporting overall accuracy, precision or recall per class.
- ✓
LIME
Why this is correct
LIME builds a locally faithful surrogate model by perturbing the instance's neighbourhood and fitting an interpretable approximation around that single prediction. This satisfies the per-prediction constraint, unlike global methods such as permutation importance or SHAP's global aggregation, which describe overall model behaviour rather than one specific outcome.
- ✗
Permutation feature importance
Why it's wrong here
Permutation feature importance measures the drop in overall model performance when a feature's values are shuffled across the dataset, yielding a global ranking rather than a per-prediction attribution. It is tempting because it quantifies feature influence, and would be correct when selecting features or reporting which inputs matter model-wide.
Go deeper
Related to this question
About these practice questions
One of 862 original AIF-C01 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 by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AIF-C01 practice question is part of Courseiva's free Amazon Web Services 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 AIF-C01 exam.