Chapter 2: Handling Data Preparation Techniques
Data is the starting point of any machine learning project, and it takes lots of work to turn data into a dataset that can be used to train a model. That work typically involves annotating datasets, running bespoke scripts to preprocess them, and saving processed versions for later use. As you can guess, doing all this work manually, or building tools to automate it, is not an exciting prospect for machine learning teams.
In this chapter, you will learn about AWS services that help you build and process data. We'll first cover Amazon SageMaker Ground Truth, a capability of Amazon SageMaker that helps you quickly build accurate training datasets. Then, we'll introduce Amazon SageMaker Data Wrangler, a new way to transform your data interactively. Next, we'll talk about Amazon SageMaker Processing, another capability that helps you run your data processing workloads, such as feature engineering, data validation, model evaluation, and model interpretation. Finally, we'll quickly discuss other AWS services that may help with data analytics: Amazon Elastic Map Reduce, AWS Glue, and Amazon Athena.
This chapter consists of the following topics:
- Labeling data with Amazon SageMaker Ground Truth
- Transforming data with Amazon SageMaker Data Wrangler
- Running batch jobs with Amazon SageMaker Processing
