AI Associate Data for AI Practice Question
A sales operations team is training an AI model to forecast quarterly revenue. They have five years of historical data, which includes a strong seasonal pattern but also a significant outlier: during the pandemic year, revenue dropped by 70% from typical values. The model trains with high accuracy on historical data but fails to predict future quarters accurately, consistently overestimating revenue. What should the data scientist do to improve forecast accuracy?
⚠ Common exam trap
Salesforce often tests the misconception that you should keep all data and adjust the model (e.g., via normalization or capacity increase) rather than removing non-representative outliers, leading candidates to pick options like C or D.
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
✓
Remove the data points corresponding to the pandemic year from the training set.
Removing the pandemic year data eliminates the extreme outlier that is causing the model to learn a distorted seasonal pattern. The 70% revenue drop is not representative of future quarters, so including it forces the model to overestimate revenue to compensate for the anomaly. By training only on typical data, the model can learn the true seasonal pattern and generalize better to future quarters.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Add a binary feature indicating whether each quarter was during the pandemic.
Why it's wrong here
Adding a feature helps the model distinguish but still includes the anomalous data; the model may still overfit to that period.
- ✓
Remove the data points corresponding to the pandemic year from the training set.
Why this is correct
Removing the outlier helps the model focus on typical patterns, improving generalization to future non-pandemic quarters.
- ✗
Normalize the entire dataset using Z-scores to reduce the impact of the outlier.
Why it's wrong here
Normalization reduces but does not eliminate the outlier's influence; the model may still be skewed.
- ✗
Include the outlier data and increase the model capacity to capture the anomaly.
Why it's wrong here
Increasing capacity may lead to overfitting on the anomaly, causing continued overestimation.
Go deeper
Related to this question
About these practice questions
Courseiva writes every AI Associate question from scratch — 753 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 Associate practice question is part of Courseiva's free Salesforce 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 Associate exam.