“Azure Machine Learning Engineering: Deploy, fine-tune, and optimize ML models using Microsoft Azure” by Sina Fakhraee is an excellent guide for anyone who wants to learn how to leverage Azure Machine Learning Services (AMLS) to build, deploy and manage machine learning models.
The book provides a comprehensive introduction to AMLS, starting with the basics of machine learning and then diving into the various features of AMLS.
What Does This Book Cover?
The book covers a wide range of topics, from data preparation to model deployment and monitoring. It provides detailed explanations of each step of the machine learning process, including data exploration, feature engineering, model selection, and hyperparameter tuning. The author also includes best practices for optimizing models, monitoring their performance, and scaling them for production use.
Chapter 1, Introducing the Azure Machine Learning Service, introduces the basic concepts of the Azure Machine Learning (AML) service. You will create an AML workspace, create a compute instance, and connect AML to VS Code for further development in later chapters.
Chapter 2, Working with Data in AMLS, covers how to work with data in AMLS. In particular, you will learn how to load data, save data as datasets, and use datasets in later development projects.
Chapter 3, Training Machine Learning Models in AMLS, shows you how to train machine learning models using AMLS experiments as well as the code-free designer. You will see how to train jobs remotely and save models to the AMLS model registry for later use.
Chapter 4, Tuning Your Models with AMLS, demonstrates how to tune hyperparameters for your machine learning models using AMLS HyperDrive.
Chapter 5, Azure Automated Machine Learning, covers how to script an AutoML job to automatically train a machine learning model.
Chapter 6, Deploying ML Models for Real-Time Inferencing, teaches you how to deploy models in the AML to support real-time inferencing.
Chapter 7, Deploying ML Models for Batch Scoring, shows you how to apply batch scoring to models using AML batch endpoints.
Chapter 8, Responsible AI, teaches you how to explain your machine learning models using AMLS and Azure Interpret.
Chapter 9, Productionizing Your Workload with MLOps, has you setting up an Azure DevOps pipeline to orchestrate model training and deployment to multiple environments.
Chapter 10, Using Deep Learning in Azure Machine Learning, demonstrates how to label image data using Azure Machine Learning’s Data Labeling feature, which we will use to train an object detection model. You will learn how to train an object detection model using AMLS AutoML and how to deploy the trained model for inferencing using AMLS.
Chapter 11, Using Distributed Training in AMLS, teaches how to perform distributed training in AMLS. In particular, you will learn how to train models in a distributed fashion using two popular deep learning frameworks, PyTorch and TensorFlow.
What sets this book apart is the author’s focus on practicality. Each chapter is accompanied by hands-on examples that demonstrate how to use AML to solve real-world problems. The author also includes tips and tricks for common challenges, such as dealing with imbalanced data and selecting the right evaluation metric.
Overall, “Azure Machine Learning Engineering” is an excellent resource for data scientists who want to learn how to use AMLS to build and deploy machine learning models. The book is well-written, easy to understand, and packed with practical examples and best practices. Whether you’re a beginner or an experienced data scientist, this book is sure to be a valuable addition to your library.
