Microsoft Certified Azure AI Fundamentals · Free Practice Question Hard
Question 26
A retail company's AI system for inventory prediction shows these characteristics:
-Misses major seasonal trends
-Performs similarly on both training and test data
-Generates overly simplified predictions
-Ignores important feature relationships
Which statement BEST describes this situation and its appropriate solution?
-
A
Overfitting requiring dimension reduction
-
B
High variance requiring regularization
-
C
High bias requiring more complex modeling and additional features
-
D
Optimal performance requiring no changes
Reveal correct answer
Correct answer: C
Explanation
Key Points on the Situation and Appropriate Solution:
High Bias Identification:
Since the model performs similarly on both training and test datasets, it suggests that the error is not due to overfitting but rather because the model is too simplistic. Overly simplified predictions and a failure to capture major trends (like seasonal variations) point to a scenario where the model is not complex enough to learn the underlying patterns.Appropriate Solution – Increasing Model Complexity and Feature Enrichment:
The best course of action is to enhance the complexity of the model. This can be achieved by:Using More Complex Modeling Techniques:
Deploying models that can capture nonlinear relationships and interactions between features—such as ensemble methods or deeper neural networks—can help in learning the complex patterns in the data.Including Additional Relevant Features:
Expanding the feature set to include variables that capture seasonal trends or other important factors can provide the model with the necessary context to improve prediction accuracy.
Why Other Options Are Less Suitable:
Approaches targeting overfitting (like regularization or dimension reduction) are designed to address high variance, not high bias.
Claiming that the model is performing optimally ignores the critical shortcomings such as missing seasonal trends and oversimplified predictions.
Further Reading and Documentation:
To better understand the bias-variance trade-off and how high bias leads to underfitting, refer to the Azure Machine Learning documentation on model evaluation and tuninglearn.microsoft.com.
Microsoft's Educational Content Around This Topic For The AI-900 Exam:
A.
Overfitting is associated with a model performing well on training data but poorly on test data due to excessive complexity. Here, the model's similar performance across both sets and its overly simplified predictions indicate underfitting rather than overfitting.
B.
High variance indicates overfitting, where a model performs well on training data but poorly on test data. In this case, the model performs similarly on both datasets, which is not characteristic of high variance.
C.
The model's failure to capture major seasonal trends, generate overly simplified predictions, and ignore important feature relationships suggests that it is underfitting, a sign of high bias. To improve performance, the team should consider using a more complex model and incorporate additional or more informative features.
D.
The model clearly misses key trends and feature interactions, indicating that it is not performing optimally. Improvements in model complexity and feature inclusion are needed to capture the underlying patterns in the data.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
