What this book covers
Chapter 1, Getting Started with Machine Learning Using Amazon SageMaker, focuses on getting our feet wet on training and deploying an ML model using Amazon SageMaker. You will perform a simplified end-to-end ML experiment using Amazon SageMaker and the Linear Learner built-in algorithm.
Chapter 2, Building and Using Your Own Algorithm Container Image, is devoted to helping us understand how model training and deployment with Amazon SageMaker works internally. You will work on creating and using your own algorithm container images and scripts to train and deploy a custom ML model in SageMaker.
Chapter 3, Using Machine Learning and Deep Learning Frameworks with Amazon SageMaker, teaches you how to define, train, and deploy your own models using several ML and deep learning frameworks with the Amazon SageMaker Python SDK. This will allow you to use any custom models you have prepared using libraries and frameworks such as TensorFlow, Keras, scikit-learn, and PyTorch, and port these to SageMaker.
Chapter 4, Preparing, Processing, and Analyzing the Data, explores the different techniques and solutions you can use to handle your different data processing and analysis requirements. You will work with recipes that make use of SageMaker Processing, Amazon Athena, and several unsupervised built-in SageMaker algorithms to perform various data preparation and processing tasks.
Chapter 5, Effectively Managing Machine Learning Experiments, provides practical solutions and examples on debugging and managing ML experiments. You will use SageMaker Debugger to detect issues in your training jobs. In addition to this, you will work with SageMaker Experiments to manage and track multiple experiments at the same time.
Chapter 6, Automated Machine Learning in Amazon SageMaker, reveals the capabilities and features of SageMaker that help us build, train, and tune ML models automatically. You will take a closer look at using AutoML in SageMaker using SageMaker Autopilot. In addition to this, you will use and configure the automatic model tuning capability to search for the optimal set of hyperparameter values for our model.
Chapter 7, Working with SageMaker Feature Store, SageMaker Clarify, and SageMaker Model Monitor, exposes a few more capabilities of SageMaker that have great integration with SageMaker Studio – SageMaker Feature Store, SageMaker Clarify, and SageMaker Model Monitor. These capabilities help data scientists and ML practitioners handle requirements that involve using online and offline feature stores, detecting bias in the data, enabling ML explainability, and monitoring a deployed model.
Chapter 8, Solving NLP, Image Classification, and Time-Series Forecasting Problems with Built-in Algorithms, is devoted to different solutions and recipes that make use of several built-in SageMaker algorithms to solve natural language processing (NLP), image classification, and time-series forecasting problems.
Chapter 9, Managing Machine Learning Workflows and Deployments, explores several intermediate solutions for real-time endpoint deployments and automated workflows. You will work on recipes that focus on deep learning model deployments for Hugging Face models, multi-model endpoint deployments, and ML workflows.
