Azure Machine learning is IDE where data scientists can develop machine learning experiments, operationalize them and integrate them into solutions using web services.
This can be accomplished with the following steps:
– Create a training experiment.
– Add your modules to it:
Click Run to run the experiment first from the bottom pane as shown below. Then:
Convert it into a predictive (scoring) experiment:
– Deploy it as a predictive web service:
You now have the ability to add, change or remove inputs and outputs in your experiment.
Click “Deploy Web Service” to compare how data flows in the experiment and in the web service
You will find new input and output nodes in the modules pane.
– You can also create a project to package different experiments and services: