Automated Machine Learning with Microsoft Azure

Automated Machine Learning with Microsoft Azure

by Dennis Michael Sawyers
Released April 2021
Publisher(s): Packt Publishing
ISBN: 9781800565319

Read it now on the O’Reilly learning platform with a 10-day free trial.

O’Reilly members get unlimited access to books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.

Start your free trial

Book description

A practical, step-by-step guide to using Microsoft's AutoML technology on the Azure Machine Learning service for developers and data scientists working with the Python programming language

Key Features

  • Create, deploy, productionalize, and scale automated machine learning solutions on Microsoft Azure
  • Improve the accuracy of your ML models through automatic data featurization and model training
  • Increase productivity in your organization by using artificial intelligence to solve common problems

Book Description

Automated Machine Learning with Microsoft Azure will teach you how to build high-performing, accurate machine learning models in record time. It will equip you with the knowledge and skills to easily harness the power of artificial intelligence and increase the productivity and profitability of your business.

Guided user interfaces (GUIs) enable both novices and seasoned data scientists to easily train and deploy machine learning solutions to production. Using a careful, step-by-step approach, this book will teach you how to use Azure AutoML with a GUI as well as the AzureML Python software development kit (SDK).

First, you'll learn how to prepare data, train models, and register them to your Azure Machine Learning workspace. You'll then discover how to take those models and use them to create both automated batch solutions using machine learning pipelines and real-time scoring solutions using Azure Kubernetes Service (AKS).

Finally, you will be able to use AutoML on your own data to not only train regression, classification, and forecasting models but also use them to solve a wide variety of business problems.

By the end of this Azure book, you'll be able to show your business partners exactly how your ML models are making predictions through automatically generated charts and graphs, earning their trust and respect.

What you will learn

  • Understand how to train classification, regression, and forecasting ML algorithms with Azure AutoML
  • Prepare data for Azure AutoML to ensure smooth model training and deployment
  • Adjust AutoML configuration settings to make your models as accurate as possible
  • Determine when to use a batch-scoring solution versus a real-time scoring solution
  • Productionalize your AutoML and discover how to quickly deliver value
  • Create real-time scoring solutions with AutoML and Azure Kubernetes Service
  • Train a large number of AutoML models at once using the AzureML Python SDK

Who this book is for

Data scientists, aspiring data scientists, machine learning engineers, or anyone interested in applying artificial intelligence or machine learning in their business will find this machine learning book useful.

You need to have beginner-level knowledge of artificial intelligence and a technical background in computer science, statistics, or information technology before getting started. Familiarity with Python will help you implement the more advanced features found in the chapters, but even data analysts and SQL experts will be able to train ML models after finishing this book.

Table of contents

  1. Automated Machine Learning with Microsoft Azure
  2. Contributors
  3. About the author
  4. About the reviewer
  5. Preface
    1. Who this book is for
    2. What this book covers
    3. To get the most out of this book
    4. Download the example code files
    5. Download the color images
    6. Conventions used
    7. Get in touch
    8. Reviews
  6. Section 1: AutoML Explained – Why, What, and How
  7. Chapter 1: Introducing AutoML
    1. Explaining data science's ROI problem
      1. Defining machine learning, data science, and AI
      2. Machine learning versus traditional software
      3. The five steps to machine learning success
      4. Putting it all together
    2. Analyzing why AI projects fail slowly
    3. Solving the ROI problem with AutoML
    4. Summary
  8. Chapter 2: Getting Started with Azure Machine Learning Service
    1. Technical requirements
    2. Creating your first AMLS workspace
      1. Creating an Azure account
      2. Creating an AMLS workspace
      3. Creating an AMLS workspace with code
      4. Navigating AML studio
    3. Building compute to run your AutoML jobs
      1. Creating a compute instance
      2. Creating a compute cluster
      3. Creating a compute cluster and compute instance with the Azure CLI
    4. Working with data in AMLS
      1. Creating a dataset using the GUI
      2. Creating a dataset using code
    5. Understanding how AutoML works on Azure
      1. Ensuring data quality with data guardrails
      2. Improving data with intelligent feature engineering
      3. Normalizing data for ML with iterative data transformation
      4. Training models quickly with iterative ML model building
      5. Getting the best results with ML model ensembling
    6. Summary
  9. Chapter 3: Training Your First AutoML Model
    1. Technical requirements
    2. Loading data into AMLS for AutoML
    3. Creating an AutoML solution
    4. Interpreting your AutoML results
      1. Understanding data guardrails
      2. Understanding model metrics
    5. Explaining your AutoML model
    6. Obtaining better AutoML performance
    7. Summary
  10. Section 2: AutoML for Regression, Classification, and Forecasting – A Step-by-Step Guide
  11. Chapter 4: Building an AutoML Regression Solution
    1. Technical requirements
    2. Preparing data for AutoML regression
      1. Setting up your Jupyter environment
      2. Preparing your data for AutoML
    3. Training an AutoML regression model
    4. Registering your trained regression model
    5. Fine-tuning your AutoML regression model
      1. Improving AutoML regression models
      2. Understanding AutoML regression algorithms
    6. Summary
  12. Chapter 5: Building an AutoML Classification Solution
    1. Technical requirements
    2. Prepping data for AutoML classification
      1. Navigating to your Jupyter environment
      2. Loading and transforming your data
    3. Training an AutoML classification model
    4. Registering your trained classification model
    5. Training an AutoML multiclass model
    6. Fine-tuning your AutoML classification model
      1. Improving AutoML classification models
      2. Understanding AutoML classification algorithms
    7. Summary
  13. Chapter 6: Building an AutoML Forecasting Solution
    1. Technical requirements
    2. Prepping data for AutoML forecasting
      1. Navigating to your Jupyter environment
      2. Loading and transforming your data
    3. Training an AutoML forecasting model
      1. Training a forecasting model with standard algorithms
      2. Training a forecasting model with Prophet and ARIMA
    4. Registering your trained forecasting model
    5. Fine-tuning your AutoML forecasting model
      1. Improving AutoML forecasting models
      2. Understanding AutoML forecasting algorithms
    6. Summary
  14. Chapter 7: Using the Many Models Solution Accelerator
    1. Technical requirements
    2. Installing the many models solution accelerator
      1. Creating a new notebook in your Jupyter environment
      2. Installing the MMSA from GitHub
    3. Prepping data for many models
      1. Prepping the sample OJ dataset
      2. Prepping a pandas dataframe
    4. Training many models simultaneously
      1. Training the sample OJ dataset
      2. Training your sample dataset with the MMSA
    5. Scoring new data for many models
      1. Scoring OJ sales data with the MMSA
      2. Scoring your sample dataset with many models
    6. Improving your many models results
    7. Summary
  15. Section 3: AutoML in Production – Automating Real-Time and Batch Scoring Solutions
  16. Chapter 8: Choosing Real-Time versus Batch Scoring
    1. Technical requirements
    2. Architecting batch scoring solutions
      1. Understanding the five-step batch scoring process
      2. Scheduling your batch scoring solution
      3. Scoring data in batches and delivering results
      4. Choosing batch over real time
    3. Architecting real-time scoring solutions
      1. Understanding the four-step real-time scoring process
      2. Training a model for real-time deployment  
      3. Delivering results in real time
      4. Knowing when to use real-time scoring
      5. Choosing real-time over batch solutions
    4. Determining batch versus real-time scoring scenarios
      1. Scenarios for real-time or batch scoring
      2. Answers for the type of solution appropriate for each scenario
    5. Summary
  17. Chapter 9: Implementing a Batch Scoring Solution
    1. Technical requirements
    2. Creating an ML pipeline
      1. Coding the first three steps of your ML scoring pipeline
      2. Creating a Python script to score data in your ML pipeline
      3. Creating and containerizing an environment
      4. Configuring and running your ML scoring pipeline
      5. Accessing your scored predictions via AML studio
    3. Creating a parallel scoring pipeline
      1. Coding the first three steps of your ML parallel scoring pipeline
      2. Creating Python scripts to score data in your ML parallel pipeline
      3. Configuring and running your ML parallel scoring pipeline
    4. Creating an AutoML training pipeline
      1. Coding the first two steps of your AutoML training pipeline
      2. Configuring your AutoML model training settings and step
      3. Creating a Python script to register your model
      4. Configuring and running your AutoML training pipeline
    5. Triggering and scheduling your ML pipelines
      1. Triggering your published pipeline from the GUI
      2. Triggering and scheduling a published pipeline through code
    6. Summary
  18. Chapter 10: Creating End-to-End AutoML Solutions
    1. Technical requirements
    2. Connecting AMLS to ADF
      1. Creating an ADF
      2. Creating a service principal and granting access
      3. Creating a linked service to connect ADF with AMLS
    3. Scheduling a machine learning pipeline in ADF
    4. Transferring data using ADF
      1. Installing a self-hosted integration runtime
      2. Creating an Azure Blob storage linked service
      3. Creating a linked service to your PC
      4. Creating an ADF pipeline to copy data
    5. Automating an end-to-end scoring solution
      1. Editing an ML pipeline to score new data
      2. Creating an ADF pipeline to run your ML pipeline
      3. Adding a trigger to your ADF pipeline
    6. Automating an end-to-end training solution
      1. Creating a pipeline to copy data into Azure
      2. Editing an ML pipeline to train with new data
      3. Adding a Machine Learning Execute Pipeline activity to your ADF pipeline
    7. Summary
  19. Chapter 11: Implementing a Real-Time Scoring Solution
    1. Technical requirements
    2. Creating real-time endpoints through the UI
      1. Creating an ACI-hosted endpoint through the UI
      2. Creating an AKS cluster through the UI
      3. Creating an AKS-hosted endpoint through the UI
    3. Creating real-time endpoints through the SDK
      1. Creating and testing a real-time endpoint with ACI through Python
      2. Creating an AKS cluster through Python
      3. Creating and testing a real-time endpoint with AKS through Python
    4. Improving performance on your AKS cluster
    5. Summary
  20. Chapter 12: Realizing Business Value with AutoML
    1. Technical requirements
    2. Architecting AutoML solutions
      1. Making key architectural decisions for AutoML solutions
      2. Architecting a batch solution
      3. Architecting a real-time solution
    3. Visualizing AutoML modeling results
      1. Visualizing the results of classification
      2. Visualizing the results of forecasting and regression
    4. Explaining AutoML results to your business
    5. Using AutoML in other Microsoft products
      1. Using AutoML within PowerBI
      2. Using AutoML within Azure Synapse Analytics
      3. Using AutoML with ML.NET
      4. Using AutoML on SQL Server, HDInsight, and Azure Databricks
    6. Realizing business value
      1. Getting the business to adopt a new, automated solution
      2. Getting the business to replace an older, automated process
      3. Getting the business to adopt a new, decision-assistance tool
      4. Getting the business to replace an old decision assistance tool
    7. Summary
    8. Why subscribe?
  21. Other Books You May Enjoy
    1. Packt is searching for authors like you
    2. Leave a review - let other readers know what you think

Product information

  • Title: Automated Machine Learning with Microsoft Azure
  • Author(s): Dennis Michael Sawyers
  • Release date: April 2021
  • Publisher(s): Packt Publishing
  • ISBN: 9781800565319