You are creating a compute target to train a machine learning experiment. The compute target must support automated machine learning, machine…

An Azure Machine Learning compute cluster is a managed-compute infrastructure specifically designed for machine learning workloads. It is the only compute target that provides the necessary scalability, management, and deep integration required to fully support all three specified features: automated machine learning (AutoML), machine learning pipelines, and the Azure Machine Learning designer. The cluster can automatically scale up from zero nodes when a job is submitted and scale down when complete, providing a cost-effective and powerful solution for training and retraining models using any of these Azure ML capabilities.

A. Azure HDInsight: This is a specialized compute target for big data processing using frameworks like Apache Spark and is not the primary or recommended compute for general use with the designer or AutoML.

C. Azure Batch: This is designed for large-scale parallel and high-performance computing (HPC) applications, which is a different use case than the integrated, iterative development common with the designer.

D. Remote VM: While a remote VM can be attached as a compute target for pipelines, it is a single, unmanaged resource that lacks the automatic scaling and management features of a compute cluster, making it less suitable.

1. Microsoft. (2023). What are compute targets in Azure Machine Learning? Azure Machine Learning Documentation. Retrieved from https://learn.microsoft.com/en-us/azure/machine-learning/concept-compute-target.

Reference Details: In the "Compute targets for training" section

the table explicitly shows that "Azure Machine Learning compute cluster" supports Azure ML designer

Automated ML

and Pipelines. The other options listed in the question do not show support for all three.

2. Microsoft. (2023). Create and manage Azure Machine Learning compute clusters. Azure Machine Learning Documentation. Retrieved from https://learn.microsoft.com/en-us/azure/machine-learning/how-to-create-attach-compute-cluster.

Reference Details: The "What is a compute cluster?" section describes it as a "managed-compute infrastructure that allows you to easily create a single or multi-node compute... The compute scales up automatically when a job is submitted." This highlights its suitability as the primary

managed training environment.