Microsoft Certified Azure Fundamentals · Free Practice Question Medium
Question 54
You're an AI specialist in a healthcare organization tasked with developing an AI model to predict disease progression based on patient data. Which of the following steps is least essential in ensuring the reliability and safety of this AI solution?
Select only one answer.
-
A
Running the model on the latest, high-end GPU to ensure maximum computational performance.
-
B
Using a diverse and representative dataset for model training to ensure it performs reliably across different demographic groups.
-
C
Implementing regular reviews and audits of the model's predictions.
-
D
Testing the model in different stages of development with a variety of data samples to ensure reliability.
Reveal correct answer
Correct answer: A
A.
While computational performance can influence the efficiency and speed of training and deploying a model, it doesn't directly ensure the reliability or safety of the AI solution.
B.
The diversity and representativeness of the training data are essential to ensure that the AI model performs reliably and safely across different patient groups.
C.
This is crucial to monitor the model's performance and to identify and correct any errors or biases, thus ensuring the reliability and safety of the AI solution.
D.
This is a common practice in machine learning and AI development and helps ensure the reliability and safety of the AI solution.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
