Build highly secure and scalable machine learning platforms to support the fast-paced adoption of machine learning solutions

Key Features

  • Explore different ML tools and frameworks to solve large-scale machine learning challenges in the cloud
  • Build an efficient data science environment for data exploration, model building, and model training
  • Learn how to implement bias detection, privacy, and explainability in ML model development

Book Description

With a highly scalable machine learning (ML) platform, organizations can quickly scale the delivery of ML products for faster business value realization, so there is a huge demand for skilled ML solutions architects in different industries. This hands-on ML book takes you through the design patterns, architectural considerations, and the latest technology that you need to know to become a successful ML solutions architect.

You'll start by understanding ML fundamentals and how ML can be applied to real-world business problems. Once you've explored some of the leading ML algorithms for solving different types of problems, the book will help you get to grips with data management and using ML libraries such as TensorFlow and PyTorch. You'll learn how to use open source technology such as Kubernetes/Kubeflow to build a data science environment and ML pipelines and then advance to building an enterprise ML architecture using Amazon Web Services (AWS) services. You'll then cover security and governance considerations, advanced ML engineering techniques, and how to apply bias detection, explainability, and privacy in ML model development. Finally, you'll get acquainted with AWS AI services and their applications in real-world use cases.

By the end of this book, you'll be able to design and build an ML platform to support common use cases and architecture patterns.

What you will learn

  • Apply ML methodologies to solve business problems
  • Design a practical enterprise ML platform architecture
  • Implement MLOps for ML workflow automation
  • Build an end-to-end data management architecture using AWS
  • Train large-scale ML models and optimize model inference latency
  • Create a business application using an AI service and a custom ML model
  • Use AWS services to detect data and model bias and explain models

Who this book is for

This book is for data scientists, data engineers, cloud architects, and machine learning enthusiasts who want to become machine learning solutions architects. Basic knowledge of the Python programming language, AWS, linear algebra, probability, and networking concepts is assumed.

Table of Contents

  1. The Machine Learning Solutions Architect Handbook
  2. Contributors
  3. About the author
  4. About the reviewers
  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. Share Your Thoughts
  6. Section 1: Solving Business Challenges with Machine Learning Solution Architecture
  7. Chapter 1: Machine Learning and Machine Learning Solutions Architecture
    1. What are AI and ML?
    2. Supervised ML
    3. Unsupervised ML
    4. Reinforcement learning
    5. ML versus traditional software
    6. ML life cycle
    7. Business understanding and ML problem framing
    8. Data understanding and data preparation
    9. Model training and evaluation
    10. Model deployment
    11. Model monitoring
    12. Business metric tracking
    13. ML challenges
    14. ML solutions architecture
    15. Business understanding and ML transformation
    16. Identification and verification of ML techniques
    17. System architecture design and implementation
    18. ML platform workflow automation
    19. Security and compliance
    20. Testing your knowledge
    21. Summary
  8. Chapter 2: Business Use Cases for Machine Learning
    1. ML use cases in financial services
    2. Capital markets front office
    3. Capital markets back office operations
    4. Risk management and fraud
    5. Insurance
    6. ML use cases in media and entertainment
    7. Content development and production
    8. Content management and discovery
    9. Content distribution and customer engagement
    10. ML use cases in healthcare and life sciences
    11. Medical imaging analysis
    12. Drug discovery
    13. Healthcare data management
    14. ML use cases in manufacturing
    15. Engineering and product design
    16. Manufacturing operations – product quality and yield
    17. Manufacturing operations – machine maintenance
    18. ML use cases in retail
    19. Product search and discovery
    20. Target marketing
    21. Sentiment analysis
    22. Product demand forecasting
    23. ML use case identification exercise
    24. Summary
  9. Section 2: The Science, Tools, and Infrastructure Platform for Machine Learning
  10. Chapter 3: Machine Learning Algorithms
    1. Technical requirements
    2. How machines learn
    3. Overview of ML algorithms
    4. Consideration for choosing ML algorithms
    5. Algorithms for classification and regression problems
    6. Algorithms for time series analysis
    7. Algorithms for recommendation
    8. Algorithms for computer vision problems
    9. Algorithms for natural language processing problems
    10. Generative model
    11. Hands-on exercise
    12. Problem statement
    13. Dataset description
    14. Setting up a Jupyter Notebook environment
    15. Running the exercise
    16. Summary
  11. Chapter 4: Data Management for Machine Learning
    1. Technical requirements
    2. Data management considerations for ML
    3. Data management architecture for ML
    4. Data storage and management
    5. Data ingestion
    6. Data cataloging
    7. Data processing
    8. Data versioning
    9. ML feature store
    10. Data serving for client consumption
    11. Authentication and authorization
    12. Data governance
    13. Hands-on exercise – data management for ML
    14. Creating a data lake using Lake Formation
    15. Creating a data ingestion pipeline
    16. Creating a Glue catalog
    17. Discovering and querying data in the data lake
    18. Creating an Amazon Glue ETL job to process data for ML
    19. Building a data pipeline using Glue workflows
    20. Summary
  12. Chapter 5: Open Source Machine Learning Libraries
    1. Technical requirements
    2. Core features of open source machine learning libraries
    3. Understanding the scikit-learn machine learning library
    4. Installing scikit-learn
    5. Core components of scikit-learn
    6. Understanding the Apache Spark ML machine learning library
    7. Installing Spark ML
    8. Core components of the Spark ML library
    9. Understanding the TensorFlow deep learning library
    10. Installing Tensorflow
    11. Core components of TensorFlow
    12. Hands-on exercise – training a TensorFlow model
    13. Understanding the PyTorch deep learning library
    14. Installing PyTorch
    15. Core components of PyTorch
    16. Hands-on exercise – building and training a PyTorch model
    17. Summary
  13. Chapter 6: Kubernetes Container Orchestration Infrastructure Management
    1. Technical requirements
    2. Introduction to containers
    3. Kubernetes overview and core concepts
    4. Networking on Kubernetes
    5. Service mesh
    6. Security and access management
    7. Network security
    8. Authentication and authorization to APIs
    9. Running ML workloads on Kubernetes
    10. Hands-on – creating a Kubernetes infrastructure on AWS
    11. Problem statement
    12. Lab instruction
    13. Summary
  14. Section 3: Technical Architecture Design and Regulatory Considerations for Enterprise ML Platforms
  15. Chapter 7: Open Source Machine Learning Platforms
    1. Technical requirements
    2. Core components of an ML platform
    3. Open source technologies for building ML platforms
    4. Using Kubeflow for data science environments
    5. Building a model training environment
    6. Registering models with a model registry
    7. Serving models using model serving services
    8. Automating ML pipeline workflows
    9. Hands-on exercise – building a data science architecture using open source technologies
    10. Part 1 – Installing Kubeflow
    11. Part 2 – tracking experiments and models, and deploying models
    12. Part 3 – Automating with an ML pipeline
    13. Summary
  16. Chapter 8: Building a Data Science Environment Using AWS ML Services
    1. Technical requirements
    2. Data science environment architecture using SageMaker
    3. SageMaker Studio
    4. SageMaker Processing
    5. SageMaker Training Service
    6. SageMaker Tuning
    7. SageMaker Experiments
    8. SageMaker Hosting
    9. Hands-on exercise – building a data science environment using AWS services
    10. Problem statement
    11. Dataset
    12. Lab instructions
    13. Summary
  17. Chapter 9: Building an Enterprise ML Architecture with AWS ML Services
    1. Technical requirements
    2. Key requirements for an enterprise ML platform
    3. Enterprise ML architecture pattern overview
    4. Model training environment
    5. Model training engine
    6. Automation support
    7. Model training life cycle management
    8. Model hosting environment deep dive
    9. Inference engine
    10. Authentication and security control
    11. Monitoring and logging
    12. Adopting MLOps for ML workflows
    13. Components of the MLOps architecture
    14. Monitoring and logging
    15. Hands-on exercise – building an MLOps pipeline on AWS
    16. Creating a CloudFormation template for the ML training pipeline
    17. Creating a CloudFormation template for the ML deployment pipeline
    18. Summary
  18. Chapter 10: Advanced ML Engineering
    1. Technical requirements
    2. Training large-scale models with distributed training
    3. Distributed model training using data parallelism
    4. Distributed model training using model parallelism
    5. Achieving low latency model inference
    6. How model inference works and opportunities for optimization
    7. Hardware acceleration
    8. Model optimization
    9. Graph and operator optimization
    10. Model compilers
    11. Inference engine optimization
    12. Hands-on lab – running distributed model training with PyTorch
    13. Modifying the training script
    14. Modifying and running the launcher notebook
    15. Summary
  19. Chapter 11: ML Governance, Bias, Explainability, and Privacy
    1. Technical requirements
    2. What is ML governance and why is it needed?
    3. The regulatory landscape around model risk management
    4. Common causes of ML model risks
    5. Understanding the ML governance framework
    6. Understanding ML bias and explainability
    7. Bias detection and mitigation
    8. ML explainability techniques
    9. Designing an ML platform for governance
    10. Data and model documentation
    11. Model inventory
    12. Model monitoring
    13. Change management control
    14. Lineage and reproducibility
    15. Observability and auditing
    16. Security and privacy-preserving ML
    17. Hands-on lab – detecting bias, model explainability, and training privacy-preserving models
    18. Overview of the scenario
    19. Detecting bias in the training dataset
    20. Explaining feature importance for the trained model
    21. Training privacy-preserving models
  20. Chapter 12: Building ML Solutions with AWS AI Services
    1. Technical requirements
    2. What are AI services?
    3. Overview of AWS AI services
    4. Amazon Comprehend
    5. Amazon Textract
    6. Amazon Rekognition
    7. Amazon Transcribe
    8. Amazon Personalize
    9. Amazon Lex
    10. Amazon Kendra
    11. Evaluating AWS AI services for ML use cases
    12. Building intelligent solutions with AI services
    13. Automating loan document verification and data extraction
    14. Media processing and analysis workflow
    15. E-commerce product recommendation
    16. Customer self-service automation with intelligent search
    17. Designing an MLOps architecture for AI services
    18. AWS account setup strategy for AI services and MLOps
    19. Code promotion across environments
    20. Monitoring operational metrics for AI services
    21. Hands-on lab – running ML tasks using AI services
    22. Summary
    23. Why subscribe?
  21. Other Books You May Enjoy
    1. Packt is searching for authors like you
    2. Share Your Thoughts