To use Azure Machine Learning

• Create an Azure Machine Learning workspace
• Create compute resources

Regression - Predict numeric label based on item's features
Classification - Predict category or class
Clustering - Group similar items into clusters(based on features)

Explore data

• Create a dataset (data for model training and other operations)

Train a machine learning model

• Multiple pre-processing techniques
• Model-training algorithms

Deploy a model as a service

• Azure Container Instances (ACI)
• Azure Kubernetes Service (AKS) cluster - recommended for production scenarios

Test the deployed service