Answer-first summary for fast verification
Answer: Create a Vertex AI Model Monitoring job. Enable feature attribution skew and drift detection for your model.
The correct answer is D. Given the scenario where you cannot move your data to the cloud due to security concerns and you need to detect changes in model performance over time, Vertex AI Model Monitoring is the appropriate tool. Specifically, enabling feature attribution skew and drift detection allows you to monitor changes in the distribution of input features, which can indicate shifts in the underlying data distribution and subsequently impact model performance. Unlike training-serving skew detection, which requires access to training data, feature attribution skew and drift detection can be done with the feature attributions calculated during inference, thus respecting the data security constraints.
Author: LeetQuiz Editorial Team
Ultimate access to all questions.
No comments yet.
You have developed a deep learning-based image classification model using on-premises data and plan to deploy it to production using Vertex AI. However, due to security policies, you cannot migrate your data to the cloud. You are concerned that the input data distribution might change over time, which could impact the model's performance. How can you detect such performance changes in production effectively using Vertex AI?
A
Use Vertex Explainable AI for model explainability. Configure feature-based explanations.
B
Use Vertex Explainable AI for model explainability. Configure example-based explanations.
C
Create a Vertex AI Model Monitoring job. Enable training-serving skew detection for your model.
D
Create a Vertex AI Model Monitoring job. Enable feature attribution skew and drift detection for your model.