Create Compute Instance

This page describes the steps required to create a compute instance for training and evaluating machine learning algorithms in Azure Machine Learning Studio.

A compute instance will provide the necessary computing resources for machine learning training and testing. While there is also an option to create a compute cluster, a low-end compute instance will be created in this tutorial.

1. Click "Settings" in top right of the "Designer"

2. Select the compute type. There are several options available but select "Compute instance" for this experiment.

3. Click "Compute instance"

4. Click "Create Azure ML compute instance"

5. Enter Compute name as "mlcomputeinstance"

6. Choose "Select from all options" to see more compute options.

7. Click on "Cost" to order all compute options by cost.

8. Select a suitable compute instance.

9. Click "Create".

10 Select the created compute instance "STANDARD_DS1_V2" in the pipeline once the instance is ready.

11. Continue with Create an ML Pipeline.