Automated Machine Learning with Microsoft Azure E-Book

Dennis Michael Sawyers

• Herausgeber: Packt Publishing
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 390

Veröffentlichungsjahr: 2021

Beschreibung

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.

Leseprobe

Automated Machine Learning with Microsoft Azure

Build highly accurate and scalable end-to-end AI solutions with Azure AutoML

Dennis Michael Sawyers

BIRMINGHAM—MUMBAI

Automated Machine Learning with Microsoft Azure

Copyright © 2021 Packt Publishing

All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without the prior written permission of the publisher, except in the case of brief quotations embedded in critical articles or reviews.

Every effort has been made in the preparation of this book to ensure the accuracy of the information presented. However, the information contained in this book is sold without warranty, either express or implied. Neither the author, nor Packt Publishing or its dealers and distributors, will be held liable for any damages caused or alleged to have been caused directly or indirectly by this book.

Packt Publishing has endeavored to provide trademark information about all of the companies and products mentioned in this book by the appropriate use of capitals. However, Packt Publishing cannot guarantee the accuracy of this information.

Group Product Manager: Kunal Parikh

Publishing Product Manager: Ali Abidi

Senior Editor: David Sugarman

Content Development Editor: Tazeen Shaikh

Technical Editor: Sonam Pandey

Copy Editor: Safis Editing

Project Coordinator: Aparna Ravikumar Nair

Proofreader: Safis Editing

Indexer: Manju Arasan

Production Designer: Vijay Kamble

First published: April 2021

Production reference: 1260321

Published by Packt Publishing Ltd.

Livery Place

35 Livery Street

Birmingham

B3 2PB, UK.

ISBN 978-1-80056-531-9

www.packt.com

To my wife, Kyoko Sawyers, who has always been by my side and supported me through many long evenings, and to my daughter, Sophia Rose, who was born halfway through the writing of this book.

– Dennis Sawyers

Contributors

About the author

Dennis Michael Sawyers is a senior cloud solutions architect (CSA) at Microsoft, specializing in data and AI. In his role as a CSA, he helps Fortune 500 companies leverage Microsoft Azure cloud technology to build top-class machine learning and AI solutions. Prior to his role at Microsoft, he was a data scientist at Ford Motor Company in Global Data Insight and Analytics (GDIA) and a researcher in anomaly detection at the highly regarded Carnegie Mellon Auton Lab. He received a master's degree in data analytics from Carnegie Mellon's Heinz College and a bachelor's degree from the University of Michigan. More than anything, Dennis is passionate about democratizing AI solutions through automated machine learning technology.

I want to thank the people who have been close to me and supported me, especially my wife, Kyoko, for encouraging me to finish this book, Rick Durham and Sam Istephan, for teaching me Azure Machine Learning, and Sabina Cartacio, Aniththa Umamahesan, and Deepti Mokkapati from the Microsoft Azure product team for helping me learn the ins and outs of AutoML.

About the reviewer

Marek Chmel is a senior CSA at Microsoft, specializing in data and AI. He is a speaker and trainer with more than 15 years' experience. He has been a Data Platform MVP since 2012. He has earned numerous certifications, including Azure Architect, Data Engineer and Scientist Associate, Certified Ethical Hacker, and several eLearnSecurity certifications. Marek earned his master's degree in business and informatics from Nottingham Trent University. He started his career as a trainer for Microsoft Server courses and later worked as SharePoint team lead and principal database administrator. He has authored two books, Hands-On Data Science with SQL Server 2017 and SQL Server 2017 Administrator's Guide.

Table of Contents

Preface

Section 1: AutoML Explained – Why, What, and How

Chapter 1: Introducing AutoML

Explaining data science's ROI problem

Defining machine learning, data science, and AI

Machine learning versus traditional software

The five steps to machine learning success

Putting it all together

Analyzing why AI projects fail slowly

Solving the ROI problem with AutoML

Summary

Chapter 2: Getting Started with Azure Machine Learning Service

Technical requirements

Creating your first AMLS workspace

Creating an Azure account

Creating an AMLS workspace

Creating an AMLS workspace with code

Navigating AML studio

Building compute to run your AutoML jobs

Creating a compute instance

Creating a compute cluster

Creating a compute cluster and compute instance with the Azure CLI

Working with data in AMLS

Creating a dataset using the GUI

Creating a dataset using code

Understanding how AutoML works on Azure

Ensuring data quality with data guardrails

Improving data with intelligent feature engineering

Normalizing data for ML with iterative data transformation

Training models quickly with iterative ML model building

Getting the best results with ML model ensembling

Summary

Chapter 3: Training Your First AutoML Model

Technical requirements

Loading data into AMLS for AutoML

Creating an AutoML solution

Interpreting your AutoML results

Understanding data guardrails

Understanding model metrics

Explaining your AutoML model

Obtaining better AutoML performance

Summary

Section 2: AutoML for Regression, Classification, and Forecasting – A Step-by-Step Guide

Chapter 4: Building an AutoML Regression Solution

Technical requirements

Preparing data for AutoML regression

Setting up your Jupyter environment

Preparing your data for AutoML

Training an AutoML regression model

Registering your trained regression model

Fine-tuning your AutoML regression model

Improving AutoML regression models

Understanding AutoML regression algorithms

Summary

Chapter 5: Building an AutoML Classification Solution

Technical requirements

Prepping data for AutoML classification

Navigating to your Jupyter environment

Loading and transforming your data

Training an AutoML classification model

Registering your trained classification model

Training an AutoML multiclass model

Fine-tuning your AutoML classification model

Improving AutoML classification models

Understanding AutoML classification algorithms

Summary

Chapter 6: Building an AutoML Forecasting Solution

Technical requirements

Prepping data for AutoML forecasting

Navigating to your Jupyter environment

Loading and transforming your data

Training an AutoML forecasting model

Training a forecasting model with standard algorithms

Training a forecasting model with Prophet and ARIMA

Registering your trained forecasting model

Fine-tuning your AutoML forecasting model

Improving AutoML forecasting models

Understanding AutoML forecasting algorithms

Summary

Chapter 7: Using the Many Models Solution Accelerator

Technical requirements

Installing the many models solution accelerator

Creating a new notebook in your Jupyter environment

Installing the MMSA from GitHub

Prepping data for many models

Prepping the sample OJ dataset

Prepping a pandas dataframe

Training many models simultaneously

Training the sample OJ dataset

Training your sample dataset with the MMSA

Scoring new data for many models

Scoring OJ sales data with the MMSA

Scoring your sample dataset with many models

Improving your many models results

Summary

Section 3: AutoML in Production – Automating Real-Time and Batch Scoring Solutions

Chapter 8: Choosing Real-Time versus Batch Scoring

Technical requirements

Architecting batch scoring solutions

Understanding the five-step batch scoring process

Scheduling your batch scoring solution

Scoring data in batches and delivering results

Choosing batch over real time

Architecting real-time scoring solutions

Understanding the four-step real-time scoring process

Training a model for real-time deployment  

Delivering results in real time

Knowing when to use real-time scoring

Choosing real-time over batch solutions

Determining batch versus real-time scoring scenarios

Scenarios for real-time or batch scoring

Answers for the type of solution appropriate for each scenario

Summary

Chapter 9: Implementing a Batch Scoring Solution

Technical requirements

Creating an ML pipeline

Coding the first three steps of your ML scoring pipeline

Creating a Python script to score data in your ML pipeline

Creating and containerizing an environment

Configuring and running your ML scoring pipeline

Accessing your scored predictions via AML studio

Creating a parallel scoring pipeline

Coding the first three steps of your ML parallel scoring pipeline

Creating Python scripts to score data in your ML parallel pipeline

Configuring and running your ML parallel scoring pipeline

Creating an AutoML training pipeline

Coding the first two steps of your AutoML training pipeline

Configuring your AutoML model training settings and step

Creating a Python script to register your model

Configuring and running your AutoML training pipeline

Triggering and scheduling your ML pipelines

Triggering your published pipeline from the GUI

Triggering and scheduling a published pipeline through code

Summary

Chapter 10: Creating End-to-End AutoML Solutions

Technical requirements

Connecting AMLS to ADF

Creating an ADF

Creating a service principal and granting access

Creating a linked service to connect ADF with AMLS

Scheduling a machine learning pipeline in ADF

Transferring data using ADF

Installing a self-hosted integration runtime

Creating an Azure Blob storage linked service

Creating a linked service to your PC

Creating an ADF pipeline to copy data

Automating an end-to-end scoring solution

Editing an ML pipeline to score new data

Creating an ADF pipeline to run your ML pipeline

Adding a trigger to your ADF pipeline

Automating an end-to-end training solution

Creating a pipeline to copy data into Azure

Editing an ML pipeline to train with new data

Adding a Machine Learning Execute Pipeline activity to your ADF pipeline

Summary

Chapter 11: Implementing a Real-Time Scoring Solution

Technical requirements

Creating real-time endpoints through the UI

Creating an ACI-hosted endpoint through the UI

Creating an AKS cluster through the UI

Creating an AKS-hosted endpoint through the UI

Creating real-time endpoints through the SDK

Creating and testing a real-time endpoint with ACI through Python

Creating an AKS cluster through Python

Creating and testing a real-time endpoint with AKS through Python

Improving performance on your AKS cluster

Summary

Chapter 12: Realizing Business Value with AutoML

Technical requirements

Architecting AutoML solutions

Making key architectural decisions for AutoML solutions

Architecting a batch solution

Architecting a real-time solution

Visualizing AutoML modeling results

Visualizing the results of classification

Visualizing the results of forecasting and regression

Explaining AutoML results to your business

Using AutoML in other Microsoft products

Using AutoML within PowerBI

Using AutoML within Azure Synapse Analytics

Using AutoML with ML.NET

Using AutoML on SQL Server, HDInsight, and Azure Databricks

Realizing business value

Getting the business to adopt a new, automated solution

Getting the business to replace an older, automated process

Getting the business to adopt a new, decision-assistance tool

Getting the business to replace an old decision assistance tool

Summary

Why subscribe?

Other Books You May Enjoy

Section 1: AutoML Explained – Why, What, and How

In this first part, you will understand why you should use AutoML and how it solves common industry problems. You will also build an AutoML solution through a UI. 

This section comprises the following chapters:

Chapter 1, Introducing AutoMLChapter 2, Getting Started with Azure Machine Learning ServiceChapter 3, Training Your First AutoML Model