Describe Azure Machine Learning Capabilities AI-900 Practice Questions (Page 8)

Questions 36–40

1. 36foundation · easy

   Which Azure Machine Learning compute option is designed to automatically scale to handle large-scale parallel training jobs?

   • ACompute instance
   • BAzure Databricks
   • CCompute cluster
   • DLocal compute
   Reveal Solution Discussion Select an answer first
2. 37expert · hard

   A data scientist wants to use automated ML but has a dataset with missing values and outliers. They need to preprocess the data before training. What should they do?

   • AManually clean the data in a separate script
   • BSkip preprocessing and train directly
   • CLet automated ML handle preprocessing automatically
   • DUse Azure Data Factory to clean the data
   Reveal Solution Discussion Select an answer first
4. 38application · medium

   A company wants to deploy a machine learning model for batch scoring on a large dataset every night. The model is registered in Azure Machine Learning. Which deployment approach should they use?

   • ADeploy the model to an AKS endpoint and send each record individually.
   • BCreate a batch endpoint and submit a batch scoring job using a compute cluster.
   • CUse Azure Container Instances to run a one-time scoring script.
   • DUse Azure Functions to process each record as it arrives.
   Reveal Solution Discussion Select an answer first
5. 39application · medium

   A data science team is experimenting with different models and needs an interactive environment with Jupyter notebooks and a GPU. They also want to use Automated ML for some experiments. Which compute option should they use?

   • AA compute instance with a GPU and pre-configured Jupyter notebooks.
   • BA low-priority compute cluster with GPU.
   • CAn Azure Databricks cluster with GPU and notebook support.
   • DA local VM with GPU and Jupyter installed.
   Reveal Solution Discussion Select an answer first
6. 40application · medium

   A company has a trained regression model registered in Azure Machine Learning. They want to deploy it as a real-time web service that can handle variable traffic and scale automatically. Which deployment target should they choose?

   • AA local Docker container on a single VM.
   • BAzure Container Instances (ACI) with a low CPU allocation.
   • CAzure Kubernetes Service (AKS) with autoscaling enabled.
   • DAn Azure Function with a custom handler.
   Reveal Solution Discussion Select an answer first