AI0-001 AI Implementation and Operations Practice Question
A data scientist is deploying a machine learning model to production. The model was trained on an imbalanced dataset. Which technique should be used during deployment to mitigate bias without retraining the model?
⚠ Common exam trap
CompTIA often tests the distinction between techniques applied during training versus deployment, and the trap here is that candidates mistakenly choose SMOTE or ensemble methods, which require retraining, instead of recognizing that threshold adjustment is a valid post-deployment bias mitigation strategy.
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 post-processing calibration to adjust decision thresholds
Post-processing calibration adjusts the decision threshold of the model to account for the class imbalance present in the training data. This technique modifies the output probabilities or classification boundary without requiring access to the original training data or retraining the model, making it suitable for deployment scenarios where the model is already fixed.
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 post-processing calibration to adjust decision thresholds
Why this is correct
Post-processing calibration adjusts decision thresholds after inference, shifting the operating point to equalise outcomes across groups. Because it modifies predictions rather than learned weights, it mitigates bias from the imbalanced training set without retraining the model.
- ✗
Use an ensemble of models trained on balanced subsets
Why it's wrong here
Training an ensemble of balanced-subset models requires fitting new estimators, which is retraining, and the stem forbids that. It tempts because balanced bagging genuinely reduces class-imbalance bias at training time, so it would be the right answer if the constraint were removed and the model could be rebuilt.
- ✗
Rebalance the dataset using SMOTE before inference
Why it's wrong here
SMOTE synthesises minority-class samples by interpolating between existing points, which alters training data and requires refitting; applying it before inference does not change the already-fitted model's learned bias. It tempts because SMOTE is the standard remedy for imbalanced training sets, correct when retraining is permitted.
- ✗
Remove sensitive features from the input data
Why it's wrong here
Dropping sensitive features does not correct bias learned from the imbalanced label distribution, and it changes the input schema the deployed model expects, causing inference failures. It tempts because feature removal is a recognised fairness technique, correct when the bias stems from protected attributes rather than class imbalance.
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 by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
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.