The Machine Learning Solutions Architect Handbook - 2nd Edition (Early Access) David Ping PDF Download
The Machine Learning Solutions Architect Handbook - 2nd Edition (Early Access) David Ping PDF Download
https://ebookmeta.com/product/the-machine-learning-solutions-
architect-handbook-2nd-edition-early-access-david-ping/
https://ebookmeta.com/product/a-first-course-in-machine-
learning-2nd-edition-solutions-manual-simon-rogers/
https://ebookmeta.com/product/machine-learning-with-python-
cookbook-practical-solutions-from-preprocessing-to-deep-
learning-2nd-ed-release-5-2nd-edition-chris-albon/
https://ebookmeta.com/product/enumerative-combinatorics-
volume-2-second-edition-richard-p-stanley-3/
The Affective Dynamics of Mass Protests 1st Edition
Ayata Bilgin
https://ebookmeta.com/product/the-affective-dynamics-of-mass-
protests-1st-edition-ayata-bilgin/
https://ebookmeta.com/product/depictions-of-children-and-the-
adult-s-journey-in-the-arts-1st-edition-margarita-georgieva/
https://ebookmeta.com/product/origin-of-the-german-
trauerspiel-1st-edition-walter-benjamin/
https://ebookmeta.com/product/a-textbook-of-internal-combustion-
engines-including-air-compressors-and-gas-turbines-and-jet-
propulsion-s-i-units-third-edition-edition-r-k-rajput/
Skye O Malley Bertrice Small
https://ebookmeta.com/product/skye-o-malley-bertrice-small/
The Machine Learning Solutions
Architect Handbook
Copyright © 2023 Packt Publishing
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, and its dealers and
distributors will be held liable for any damages caused or alleged to
be caused directly or indirectly by this book.
Livery Place
35 Livery Street
Birmingham
B3 2PB, UK
ISBN: 978-1-80512-250-0
www.packt.com
Table of Contents
1. The Machine Learning Solutions Architect Handbook, Second
Edition: A comprehensive guide to AI use cases, enterprise ML
platform and solutions design on AWS, AI risk management,
and generative AI
2. 1 Machine Learning and Machine Learning Solutions Architecture
I. Join our book community on Discord
II. What are AI and ML?
i. Supervised ML
ii. Unsupervised ML
iii. Reinforcement learning
III. ML versus traditional software
IV. ML life cycle
i. Business understanding and ML problem framing
ii. Data understanding and data preparation
iii. Model training and evaluation
iv. Model deployment
v. Model monitoring
vi. Business metric tracking
V. ML challenges
VI. ML solutions architecture
i. Business understanding and ML transformation
ii. Identification and verification of ML techniques
iii. System architecture design and implementation
iv. ML platform workflow automation
v. Security and compliance
VII. Testing your knowledge
VIII. Summary
3. 2 Business Use Cases for Machine Learning
I. Join our book community on Discord
II. ML use cases in financial services
i. Capital markets front office
ii. Capital markets back office operations
iii. Risk management and fraud
iv. Insurance
III. ML use cases in media and entertainment
i. Content development and production
ii. Content management and discovery
iii. Content distribution and customer engagement
IV. ML use cases in healthcare and life sciences
i. Medical imaging analysis
ii. Drug discovery
iii. Healthcare data management
V. ML use cases in manufacturing
i. Engineering and product design
ii. Manufacturing operations – product quality and yield
iii. Manufacturing operations – machine maintenance
VI. ML use cases in retail
i. Product search and discovery
ii. Target marketing
iii. Sentiment analysis
iv. Product demand forecasting
VII. ML use cases in automotive
i. Autonomous vehicle
ii. Advanced driver assistance systems (ADAS)
VIII. Summary
4. 3 Machine Learning Algorithms
I. Join our book community on Discord
II. Technical requirements
III. How machines learn
IV. Overview of ML algorithms
i. Consideration for choosing ML algorithms
ii. Algorithms for classification and regression problems
iii. Algorithms for clustering
iv. Algorithms for time series analysis
v. Algorithms for recommendation
vi. Algorithms for computer vision problems
vii. Algorithms for natural language processing (NLP)
problems
viii. Generative AI algorithms
V. Hands-on exercise
i. Problem statement
ii. Dataset description
iii. Setting up a Jupyter Notebook environment
iv. Running the exercise
VI. Summary
5. 4 Data Management for Machine Learning
I. Join our book community on Discord
II. Technical requirements
III. Data management considerations for ML
IV. Data management architecture for ML
V. Data storage and management
i. AWS Lake Formation
VI. Data ingestion
i. Kinesis Firehose
ii. AWS Glue
iii. AWS Lambda
VII. Data cataloging
i. AWS Glue Catalog
ii. Custom data catalog solution
VIII. Data processing
i. AWS Glue ETL
ii. Amazon Elastic Map Reduce (EMR)
iii. AWS Lambda
IX. ML data versioning
i. S3 partitions
ii. Versioned S3 buckets
iii. Purpose-built data version tools
X. ML feature store
XI. Data serving for client consumption
i. Consumption via API
ii. Consumption via data copy
XII. Special databases for ML
i. Vector Database
ii. Graph Databases
XIII. Data pipeline
i. AWS Glue workflows
ii. AWS Managed Airflow
XIV. Authentication and authorization
XV. Data governance
i. Data Lineage
ii. Other data governance measures
XVI. Hands-on exercise – data management for ML
i. Creating a data lake using Lake Formation
ii. Creating a data ingestion pipeline
iii. Creating a Glue catalog
iv. Discovering and querying data in the data lake
v. Creating an Amazon Glue ETL job to process data for
ML
vi. Building a data pipeline using Glue workflows
XVII. Summary
6. 5 Open Source Machine Learning Libraries
I. Join our book community on Discord
II. Technical requirements
III. Core features of open source machine learning libraries
IV. Understanding the scikit-learn machine learning library
i. Installing scikit-learn
ii. Core components of scikit-learn
V. Understanding the Apache Spark ML machine learning
library
i. Installing Spark ML
ii. Core components of the Spark ML library
VI. Understanding the TensorFlow deep learning library
i. Installing Tensorflow
ii. Core components of TensorFlow
VII. Hands-on exercise – training a TensorFlow model
VIII. Understanding the PyTorch deep learning library
i. Installing PyTorch
ii. Core components of PyTorch
IX. Hands-on exercise – building and training a PyTorch model
X. Summary
7. 6 Kubernetes Container Orchestration Infrastructure
Management
I. Join our book community on Discord
II. Technical requirements
III. Introduction to containers
IV. Kubernetes overview and core concepts
i. Namespaces
ii. Pods
iii. Deployment
iv. Kubernetes Job
v. Kubernetes custom resources ( CRs ) and
operators
vi. Services
V. Networking on Kubernetes
i. Service mesh
VI. Security and access management
i. Network security
ii. Authentication and authorization to APIs
iii. Running ML workloads on Kubernetes
VII. Hands-on – creating a Kubernetes infrastructure on AWS
i. Problem statement
ii. Lab instruction
VIII. Summary
8. 7 Open Source Machine Learning Platforms
I. Join our book community on Discord
II. Core components of an ML platform
III. Open source technologies for building ML platforms
i. Implementing a data science environment
ii. Building a model training environment
iii. Registering models with a model registry
iv. Serving models using model serving services
v. Monitoring models in production
vi. Managing ML features
vii. Feast
viii. Automating ML pipeline workflows
IV. Building end-to-end ML platform
i. ML platform based strategy
ii. ML component based strategy
V. Summary
9. 8 Building a Data Science Environment Using AWS ML Services
I. Join our book community on Discord
II. Technical requirements
III. SageMaker overview
IV. Data science environment architecture using SageMaker
i. Onboarding SageMaker users
ii. Launching a notebook
iii. Preparing data
iv. Creating, storing, and sharing features
v. Training machine learning models
vi. Tuning machine learning models
vii. Deploying machine learning models for testing
V. Automating experimentation and model building
VI. Best practices for building data science environment
VII. Hands-on exercise – building a data science environment
using AWS services
i. Problem statement
ii. Dataset
iii. Lab instructions
VIII. Summary
10. 9 Building an Enterprise ML Architecture with AWS ML Services
I. Join our book community on Discord
II. Technical requirements
III. The personas of ML platforms and their requirement
i. ML platform builder
ii. Platform user and operator
iii. Common workflow of an ML initiative
iv. Platform capability requirements
IV. Key requirements for an enterprise ML platform
V. Enterprise ML architecture pattern overview
VI. Model training environment
i. Model training engine using SageMaker
ii. Automation support
iii. Model training life cycle management
VII. Model hosting environment deep dive
i. Inference engine
ii. Authentication and security control
iii. Monitoring and logging
VIII. Adopting MLOps for ML workflows
i. Components of the MLOps architecture
ii. Monitoring and logging
iii. Best practices in building and operating ML platform
IX. Hands-on exercise – building an MLOps pipeline on AWS
i. Creating a CloudFormation template for the ML training
pipeline
ii. Creating a CloudFormation template for the ML
deployment pipeline
X. Summary
11. 12 AI Risk Management
I. Join our book community on Discord
II. Understanding AI risk scenarios
i. Bias and discrimination
ii. Misinformation and misinterpretation
iii. Lack of interpretability
iv. Unintended consequences
v. Adversarial Attacks
vi. Privacy violation and sensitive data exposure
vii. Third party risk
viii. Model testing risk
III. The regulatory landscape around AI risk management
IV. Understanding AI risk management
i. Placing trust in AI systems
V. Creating risk management framework for trustworthy AI
i. Governance oversight Principals
ii. AI risk management framework
VI. Applying risk management across AI lifecycle
i. Business problem identification and definition
ii. Data acquisition and management
iii. Experimentation and model development
iv. AI system deployment and operations
VII. Designing ML platform for governance and risk
considerations
i. Data and model documentation
ii. Lineage and reproducibility
iii. Observability and auditing
iv. Scalability and performance
v. Data quality
VIII. Summary
12. 13 Bias, Explainability, Privacy, and Adversarial attacks
I. Join our book community on Discord
II. Understanding bias
i. Bias detection and mitigation
III. Understanding ML explainability
IV. Understanding security and privacy preserving ML
V. Understanding Adversarial Attacks
i. Evasion attacks
ii. Data poisoning attacks
iii. Clean-label Backdoor Attack
iv. Model extraction attack
VI. Defense against adversarial attacks
i. Robustness-based methods
ii. Detector-based method
iii. Open-source tools for adversarial attacks and defenses
VII. Hands-on Lab – Detecting bias, explaining model, and
training privacy-preserving models
i. Overview of the scenario
ii. Detecting bias in the training dataset
iii. Explaining feature importance for trained model
iv. Training privacy preserving models
VIII. Summary
13. 14 Progressing the ML journey
I. Join our book community on Discord
II. Understanding ML journey and AI capabilities
i. ML adoption stages
ii. AI/ML maturity and assessment
iii. AI/ML operating models
III. Solving ML Journey challenges
i. Developing AI vision and strategy
ii. Getting started with first AI/ML initiative
iii. Solving scaling challenge with AI/ML adoption
IV. Summary
The Machine Learning Solutions
Architect Handbook, Second
Edition: A comprehensive guide to
AI use cases, enterprise ML platform
and solutions design on AWS, AI risk
management, and generative AI
Welcome to Packt Early Access. We’re giving you an exclusive
preview of this book before it goes on sale. It can take many months
to write a book, but our authors have cutting-edge information to
share with you today. Early Access gives you an insight into the
latest developments by making chapter drafts available. The
chapters may be a little rough around the edges right now, but our
authors will update them over time.You can dip in and out of this
book or follow along from start to finish; Early Access is designed to
be flexible. We hope you enjoy getting to know more about the
process of writing a Packt book.