AI-900 Practice Question: Describe fundamental principles of machine learning on Azure
A data scientist is training a binary classification model to detect rare equipment failures from sensor data. The dataset contains 99.5% normal operation readings and only 0.5% failure readings. The model currently predicts all readings as 'normal' and achieves 99.5% accuracy on the test set. The business requires the model to identify at least 80% of actual failures. Which data-level technique should the data scientist use to most directly address the class imbalance?
⚠ Common exam trap
Many exam-takers think high accuracy (99.5%) is always good, but in imbalanced datasets, accuracy is misleading; the question tests whether you recognize that data-level techniques like oversampling are needed to force the model to learn the minority class, not just optimize metrics or simplify the model.
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
✓
Oversample the minority class (failure examples)
Oversampling the minority class (failure examples) directly addresses the severe class imbalance by creating synthetic copies or duplicates of the rare failure instances. This balances the training dataset, allowing the model to learn patterns associated with failures rather than always predicting the majority class. With a balanced dataset, the model can be trained to meet the business requirement of identifying at least 80% of actual failures, even though overall accuracy may decrease.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Oversample the minority class (failure examples)
Why this is correct
Oversampling the minority class replicates or synthetically generates additional failure examples so the training set has a more balanced class distribution. This directly gives the model more opportunities to learn the boundary of rare failure patterns, reducing the tendency to simply predict the majority class. Techniques like SMOTE create interpolated samples rather than exact duplicates, which generally improves generalization on the minority class.
- ✗
Undersample the majority class (normal examples)
Why it's wrong here
Undersampling discards a large portion of the normal examples to match the count of failures, which balances the classes but throws away potentially informative data about what normal behavior looks like. This loss of majority-class information can blur decision boundaries and increase variance, especially when the dataset is already small. Moreover, it does not create any new failure examples, so the model may still lack enough minority patterns to learn from.
- ✗
Use precision as the optimization metric
Why it's wrong here
Using precision as the optimization metric only changes how model performance is evaluated, not the underlying data imbalance. Precision emphasizes avoiding false alarms, so the model may become conservative and miss actual failures, which is counterproductive when the goal is to detect rare failures. The minority class remains severely underrepresented in training, so the model still has little evidence to learn failure patterns regardless of the metric.
- ✗
Reduce the complexity of the model
Why it's wrong here
Reducing model complexity lowers its capacity to capture intricate patterns, and on imbalanced data this typically pushes the model toward the dominant class, worsening minority-class recall. This approach does not address the root cause, which is the scarcity of failure examples, and often leads to underfitting. A simpler model would also be less able to model the nuanced feature relationships required to separate rare failures from normal operation.
Go deeper
Related to this question
Learn chapter
Regression and Classification
Key term
Classification
Classification is a supervised machine learning technique used to predict a category or class label for new data based on patterns learned from labeled training data.
Key term
Model
In IT and AI, a model is a trained mathematical representation that learns patterns from data to make predictions or decisions.
About these practice questions
One of 985 original AI-900 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 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.