In the world of machine learning, automated training pipelines streamline the journey from data to insight. They automate various parts of the machine learning life cycle such as data ingestion, preprocessing, model training, evaluation and deployment. Amazon Web Services (“AWS”) provides various tools to develop an automated training pipeline. In this article, we will walk through setting up a basic automated training pipeline using using the classic iris dataset.
Setting the Stage: Requirements and AWS Toolkit
In this section, we will cover some high level requirements as well as a brief overview of the AWS tools we will use.
Requirements
If you choose to follow along by building your own training pipeline, you will need the following.