Before building or training any model in Azure Machine Learning Studio-including when using Automated ML (AutoML)-you must first create a Machine Learning workspace.
A workspace serves as the central environment for all machine learning assets such as datasets, compute targets, models, pipelines, and experiments. According to the AI-900 study guide and Microsoft Learn module
"Describe features and tools for machine learning in Azure," a workspace is the foundational setup required to organize and manage all ML-related resources.
The sequence typically follows these steps:
* Create a Machine Learning workspace.
* Configure compute resources (e.g., compute instance or cluster).
* Upload or register datasets.
* Use Automated ML or Designer to train models.
* Deploy and manage the trained models.
Option A (Jupyter notebook) is an optional tool for coding experiments.
Option C (Registered dataset) is created after the workspace exists.
Option D (Designer pipeline) is a visual tool used within the workspace.
Hence, B. a Machine Learning workspace is the correct answer because it is the first and mandatory step before using Automated ML or any training component in Azure Machine Learning Studio.