Implement machine learning model lifecycle and operations Questions
Practice questions for Implement machine learning model lifecycle and operations topic in Microsoft Certified: Machine Learning Operations (MLOps) Engineer Associate. 57 questions covering this domain.
A team wants Azure Machine Learning to explore candidate models automatically for a supervised learning problem. Which capability should they use?
A data scientist wants to record parameters, metrics, and artifacts for training runs by using the experiment tracking capability called out in the AI...
A model in production is still returning predictions, but the input population has changed significantly from the training data. Which monitoring conc...
A model has just finished training and must be tracked for versioned deployment and governance. What should the engineer do next?
An MLOps team wants production models to trigger action when performance or drift thresholds are exceeded. Which design best matches the study guide?
A new model deployment receives 10 percent of production traffic and starts producing worse predictions than the current version. Which response best ...
A deep learning workload exceeds the capacity of a single node and needs coordinated training across several machines. Which design best fits Azure Ma...
A team wants to keep older model versions for audit purposes but remove them from normal list views so only current versions are visible during day-to...
A batch scoring job usually runs with default settings, but one urgent run needs more instances and a different mini-batch size without changing the p...
A deployment to a managed online endpoint fails during provisioning, and the engineer needs the fastest next step to inspect what happened on the serv...
A team must score millions of files overnight and can wait for the results to land in storage. Which Azure Machine Learning feature should they use?
An application needs synchronous low-latency predictions over HTTPS for each request. Which Azure Machine Learning deployment approach is the best fit...
A practitioner wants to try multiple hyperparameter combinations and stop poor-performing trials early to save time and cost. Which capability best ma...
A machine learning engineer wants to build a reusable workflow that includes data preparation, training, validation, and registration as separate step...
A production endpoint needs a controlled release of a new model version so only part of traffic reaches it at first. Which deployment pattern should b...
A data scientist wants metrics, parameters, and the trained model logged automatically when training a scikit-learn model in Azure Machine Learning, w...
To detect data drift in production, an engineer enables Azure Machine Learning model monitoring. Which capability captures inference inputs and output...
Azure Machine Learning model monitoring is configured. Which built-in signal compares the production feature distribution to a baseline to detect inpu...
Which Azure Machine Learning command (CLI v2) creates a new managed online endpoint from a YAML file?
An engineer wants to generate fast iteration feedback by running a deployment locally on their workstation before pushing it to the cloud. Which CLI v...
Sign in to see all 57 questions
Create a free account to browse all questions — completely free during our launch phase.