AI-900 Practice Question: Describe Artificial Intelligence workloads and considerations
A data scientist is training a credit risk model and wants to use Azure Machine Learning's Responsible AI dashboard to identify if the model is biased against a certain demographic group. Which component of the dashboard should they use to evaluate this?
⚠ Common exam trap
Test-takers frequently confuse Model Interpretability (which explains why a model made a prediction) with Fairness Assessment (which evaluates bias across groups), leading them to select Option A when the question specifically asks about bias against a demographic group.
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
✓
Model Fairness Assessment
The Model Fairness Assessment component of Azure Machine Learning's Responsible AI dashboard is specifically designed to evaluate and mitigate bias in machine learning models. It allows data scientists to assess disparities in model performance across demographic groups defined by sensitive features (e.g., race, gender) using metrics like demographic parity, equal opportunity, and disparate impact. This directly addresses the question of identifying bias against a certain demographic group.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Model Interpretability
Why it's wrong here
Model Interpretability techniques, such as SHAP values, LIME, or feature importance, explain why a specific prediction was made by attributing outcomes to input features. While these insights can reveal that a model relies heavily on a proxy for a sensitive attribute, they do not systematically compare error rates or prediction consistency across predefined protected groups. For fairness auditing, we need a quantitative, group-level comparison rather than a per-instance narrative.
When this WOULD be correct
A question asking which component helps understand why a model made a specific prediction for a loan applicant, such as identifying the key factors leading to a high-risk score.
- ✓
Model Fairness Assessment
Why this is correct
Model Fairness Assessment directly evaluates the trained model's predictions across user-defined sensitive groups such as race, ethnicity, or gender. It computes fairness metrics like demographic parity, equalized odds, and disparate impact to quantify whether the model treats these groups unequally. Unlike data-focused checks, this component operates on model outputs, making it the appropriate tool for ensuring the resultant credit risk model does not encode discriminatory behavior.
- ✗
Error Analysis
Why it's wrong here
Error Analysis focuses on finding cohorts, data subsets, or 'error slices' where the model has particularly high error rates, often revealing patterns like geographic or product-type issues. Its slices are not inherently defined by protected attributes, and its goal is to improve overall accuracy or identify difficult cases. It does not enforce the specific fairness constraint of measuring bias across sensitive groups, so it is insufficient for addressing systematic demographic discrimination.
When this WOULD be correct
A question asking: 'A data scientist wants to identify which subsets of data have the highest prediction errors to improve model accuracy. Which component of the Responsible AI dashboard should they use?' would make Error Analysis the correct answer.
- ✗
Data Balance Analysis
Why it's wrong here
Data Balance Analysis inspects the training dataset to detect imbalances such as underrepresentation of a sensitive group or skewed label distributions across attributes. Although severe data imbalance can breed biased outcomes, this check occurs before model training and does not measure whether the final predictions actually favor or penalize any demographic segment. Predicting fairness requires evaluating post-training outputs, not merely examining input statistics.
When this WOULD be correct
A question asking: 'A data scientist wants to check if the training dataset has sufficient representation of all demographic groups before training a model. Which component should they use?'
Option-by-option analysis
Why each answer is right or wrong
Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The AI-900 exam frequently reuses these exact scenarios with slightly different constraints.
✓Model Fairness AssessmentCorrect answer▾
Why this is correct
Model Fairness Assessment directly evaluates the trained model's predictions across user-defined sensitive groups such as race, ethnicity, or gender. It computes fairness metrics like demographic parity, equalized odds, and disparate impact to quantify whether the model treats these groups unequally. Unlike data-focused checks, this component operates on model outputs, making it the appropriate tool for ensuring the resultant credit risk model does not encode discriminatory behavior.
✗Model InterpretabilityWrong answer — click to see why▾
Why this is wrong here
Model Interpretability explains how features influence predictions but does not evaluate bias or fairness across demographic groups.
★ When this WOULD be the correct answer
A question asking which component helps understand why a model made a specific prediction for a loan applicant, such as identifying the key factors leading to a high-risk score.
Why candidates choose this
Candidates may confuse interpretability with fairness, assuming that understanding model decisions inherently reveals bias.
✗Error AnalysisWrong answer — click to see why▾
Why this is wrong here
Error Analysis focuses on identifying regions of high error in the model's predictions, not on evaluating bias against demographic groups. The question specifically asks for bias assessment, which is the role of Model Fairness Assessment.
★ When this WOULD be the correct answer
A question asking: 'A data scientist wants to identify which subsets of data have the highest prediction errors to improve model accuracy. Which component of the Responsible AI dashboard should they use?' would make Error Analysis the correct answer.
Why candidates choose this
Candidates may confuse 'error' with 'bias', thinking that analyzing errors will reveal unfair treatment of groups, but error analysis does not directly measure fairness or demographic disparities.
✗Data Balance AnalysisWrong answer — click to see why▾
Why this is wrong here
Data Balance Analysis is used to detect imbalances in the training data (e.g., underrepresentation of a group), but the question asks about identifying bias in the model's predictions, which requires fairness assessment of the model's outputs.
★ When this WOULD be the correct answer
A question asking: 'A data scientist wants to check if the training dataset has sufficient representation of all demographic groups before training a model. Which component should they use?'
Why candidates choose this
Candidates may confuse data imbalance (a cause of bias) with model fairness (the effect), assuming that analyzing data balance directly evaluates model bias.
Analysis generated from the official AI-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”
Quick reference
RAID Level Comparison
| RAID Level | Min Disks | Fault Tolerance | Read | Write | Usable Capacity |
|---|---|---|---|---|---|
| RAID 0 | 2 | None | Excellent | Excellent | 100% |
| RAID 1 | 2 | 1 disk | Good | Moderate | 50% |
| RAID 5 | 3 | 1 disk | Good | Moderate | 67–94% |
| RAID 6 | 4 | 2 disks | Good | Lower | 50–88% |
| RAID 10 | 4 | 1 disk per mirror | Excellent | Good | 50% |
RAID is not a backup strategy — it protects against disk failure but not against accidental deletion, ransomware, or site-level events.
Go deeper
Related to this question
Learn chapter
Responsible AI Principles
Key term
Azure Machine Learning
Azure Machine Learning is a cloud service for building, training, and deploying machine learning models at scale.
Key term
Machine learning
Machine learning is a branch of artificial intelligence where computers learn patterns from data to make decisions or predictions without being explicitly programmed for every task.
About these practice questions
Courseiva writes every AI-900 question from scratch — 985 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI-900 practice question is part of Courseiva's free Microsoft 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 AI-900 exam.