Demonstration showing how to use Azure Machine Learning Studio to predict income levels
An Azure RM template is included in infrastructure/
to create the environment:
Using Azure PowerShell, do the following to provision the resources:
.\startup.ps1
Alternatively, you can perform a one-click deploy with the following button:
- Navigate to the Azure Machine Learning workspace created by the ARM template and click the link to open Azure Machine Learning Studio
- Create a new Experiment
- Build the following model in the designer:
- In the Select Columns for Dataset module properties, remove a subset of features from the dataset so you can see how important they are compared to the model that is trained with all features
- Run the experiment
- Compare the model performance by selecting the Evaluate Model's visualize option
When finished, remove the Azure resources with:
.\teardown.ps1
Thanks to the University of California Irvine Machine Learning Lab for hosting the data here.