Book Image

Learn Amazon SageMaker - Second Edition

By : Julien Simon
Book Image

Learn Amazon SageMaker - Second Edition

By: Julien Simon

Overview of this book

Amazon SageMaker enables you to quickly build, train, and deploy machine learning models at scale without managing any infrastructure. It helps you focus on the machine learning problem at hand and deploy high-quality models by eliminating the heavy lifting typically involved in each step of the ML process. This second edition will help data scientists and ML developers to explore new features such as SageMaker Data Wrangler, Pipelines, Clarify, Feature Store, and much more. You'll start by learning how to use various capabilities of SageMaker as a single toolset to solve ML challenges and progress to cover features such as AutoML, built-in algorithms and frameworks, and writing your own code and algorithms to build ML models. The book will then show you how to integrate Amazon SageMaker with popular deep learning libraries, such as TensorFlow and PyTorch, to extend the capabilities of existing models. You'll also see how automating your workflows can help you get to production faster with minimum effort and at a lower cost. Finally, you'll explore SageMaker Debugger and SageMaker Model Monitor to detect quality issues in training and production. By the end of this Amazon book, you'll be able to use Amazon SageMaker on the full spectrum of ML workflows, from experimentation, training, and monitoring to scaling, deployment, and automation.
Table of Contents (19 chapters)
1
Section 1: Introduction to Amazon SageMaker
4
Section 2: Building and Training Models
11
Section 3: Diving Deeper into Training
14
Section 4: Managing Models in Production

Chapter 3: AutoML with Amazon SageMaker Autopilot

In the previous chapter, you learned how Amazon SageMaker helps you build and prepare datasets. In a typical machine learning project, the next step would be to start experimenting with algorithms in order to find an early fit and get a sense of the predictive power you could expect from the model.

Whether you work with traditional machine learning or deep learning, three options are available when it comes to selecting an algorithm:

  • Write your own, or customize an existing one. This only makes sense if you have strong skills in statistics and computer science, if you're quite sure that you can do better than well-tuned, off-the-shelf algorithms, and if you're given enough time to work on the project. Let's face it, these conditions are rarely met.
  • Use a built-in algorithm implemented in one of your favorite libraries, such as linear regression or XGBoost. For deep learning problems, this includes pre-trained...