Prepare for the AWS Machine Learning Engineer exam smarter and faster and get job-ready with this efficient and authoritative resource
In AWS Certified Machine Learning Engineer Study Guide: Associate (MLA-C01) Exam, veteran AWS Practice Director at Trace3—a leading IT consultancy offering AI, data, cloud and cybersecurity solutions for clients across industries—Dario Cabianca delivers a practical and up-to-date roadmap to preparing for the MLA-C01 exam. You'll learn the skills you need to succeed on the exam as well as those you need to hit the ground running at your first AI-related tech job.
You'll learn how to prepare data for machine learning models on Amazon Web Services, build, train, refine models, evaluate model performance, deploy and secure your machine learning applications against bad actors.
Inside the book:
- Complimentary access to the Sybex online test bank, which includes an assessment test, chapter review questions, practice exam, flashcards, and a searchable key term glossary
- Strategies for selecting and justifying an appropriate machine learning approach for specific business problems and identifying the most efficient AWS solutions for those problems
- Practical techniques you can implement immediately in an artificial intelligence and machine learning (AI/ML) development or data science role
Perfect for everyone preparing for the AWS Certified Machine Learning Engineer -- Associate exam, AWS Certified Machine Learning Engineer Study Guide is also an invaluable resource for those preparing for their first role in AI or data science, as well as junior-level practicing professionals seeking to review the fundamentals with a convenient desk reference.
Table of Contents:
Contents
Chapter 1Introduction to Machine Learning1
Understanding Artificial Intelligence2
Data, Information, Knowledge3
Data3
Information4
Knowledge5
Understanding Machine Learning6
ML Lifecycle6
Define ML Problem6
Collect Data8
Process Data8
Choose Algorithm8
Train Model9
Evaluate Model9
Deploy Model9
Derive Inference11
Monitor Model11
ML Concepts11
Features11
Target Variable12
Optimization Problem12
Objective Function13
ML Algorithms vs. ML Models13
Differences Between ML and AI14
Understanding Deep Learning16
Introduction to Neural Networks16
Structure of a Neural Network16
Neuron16
Input Layer18
Hidden Layers18
Output Layer18
How Neural Networks Work18
Neural Networks Types19
Artificial Neural Networks20
Deep Neural Networks20
Convolutional Neural Networks20
Recurrent Neural Networks20
Differences Between DL and ML21
Case Studies21
Case Study 1: Mobileye’s Autonomous Driving Technology21
Case Study 2: Leidos’ Healthcare ML Applications21
Summary22
Exam Essentials23
Review Questions24
Chapter 2Data Ingestion and Storage27
Introducing Ingestion and Storage28
Ingesting and Storing Data28
Data Formats and Ingestion Techniques31
Choosing AWS Ingestion Services34
Amazon Data Firehose35
Amazon Kinesis Data Streams35
Amazon Managed Streaming for Apache Kafka (MSK)36
Amazon Managed Service for Apache Flink38
AWS DataSync39
AWS Glue40
Choosing AWS Storage Services41
Amazon Simple Storage Service (S3)42
Amazon Elastic File System (EFS)45
Amazon FSx for Lustre47
Amazon FSx for NetApp ONTAP49
Amazon FSx for Windows File Server50
Amazon FSx for OpenZFS51
Amazon Elastic Block Storage (EBS)51
Amazon Relational Database Service (RDS)52
Amazon DynamoDB52
Troubleshooting53
Summary54
Exam Essentials55
Review Questions57
Chapter 4Model Selection61
Understanding AWS AI Services63
Vision64
Amazon Rekognition64
Amazon Textract65
Speech66
Amazon Polly66
Amazon Transcribe67
Language67
Amazon Translate67
Amazon Comprehend68
Chatbot69
Amazon Lex69
Recommendation70
Amazon Personalize70
Generative AI71
Amazon Bedrock71
Developing Models with Amazon SageMaker Built-in Algorithms81
Supervised ML Algorithms81
General Regression and Classification Algorithms83
Recommendation102
Forecasting104
Unsupervised ML Algorithms105
Clustering105
Dimensionality Reduction113
Topic Modeling119
Anomaly Detection121
Textual Analysis123
BlazingText124
Sequence-to-Sequence126
Image Processing127
Image Classification127
Object Detection128
Semantic Segmentation130
Criteria for Model Selection131
Summary132
Exam Essentials133
Review Questions136
Chapter 5Model Training and Evaluation141
Training143
Local Training144
Remote Training145
Distributed Training146
Monitoring Training Jobs147
Debugging Training Jobs148
Hyperparameter Tuning149
Model Parameter and Hyperparameter151
Exploring the Hyperparameter Space with Amazon SageMaker AI Automatic Model Tuning152
Evaluation Metrics154
Classification Problem Metrics154
Regression Problem Metrics160
Hyperparameter Tuning Techniques164
Manual Search164
Grid Search165
Random Search165
Bayesian Search165
Multi-algorithm Optimization166
Managing Bias and Variance Trade-Off166
Addressing Overfitting and Underfitting168
Underfitting168
Overfitting170
Regularization170
Advanced Techniques173
Model Performance Evaluation173
Performance Evaluation Methods173
K-Fold Cross-Validation174
Random Train-Test Split175
Holdout Set176
Bootstrap176
Evaluating Foundation Models177
Automatic Evaluations177
Human Evaluations177
LLM-as-a-Judge177
Programmatic Evaluations177
Knowledge Base Evaluations177
Deep Dive Model Tuning Example177
Summary185
Exam Essentials187
Review Questions190
Chapter 6Model Deployment and Orchestration193
AWS Model Deployment Services194
Deploying AI Services195
Amazon Rekognition196
Amazon Textract197
Amazon Polly197
Amazon Transcribe198
Amazon Comprehend198
Amazon Lex199
Amazon Personalize199
Amazon Bedrock200
Deploying Your Model201
Infrastructure Selection Considerations202
Managed Model Deployments203
Unmanaged Model Deployments211
Optimizing ML Models for Edge Devices216
Advanced Model Deployment Techniques218
Autoscaling Endpoints218
Deployment and Testing Strategies221
Blue/Green Deployment221
Orchestrating ML Workflows227
Introducing Amazon SageMaker Pipelines228
Code Repository and Version Control228
Introducing Amazon SageMaker Model Registry229
CI/CD230
MLOps Orchestration230
AWS Step Functions231
Amazon Managed Workflows for Apache Airflow232
Choosing an Orchestration Tool232
Automating Model Building and Deployment233
Define the Workflow Steps234
Create and Configure Pipeline Steps234
Define the Pipeline237
Set Up Triggers and Schedules237
Execute the Pipeline238
Key Considerations238
Deep-Dive Model Deployment Example238
Summary247
Exam Essentials248
Review Questions250
Chapter 7Model Monitoring and Cost Optimization253
Monitoring Model Inference255
Drifts in Models256
Techniques to Monitor Data Quality and Model Performance257
Monitoring Workflow259
Design Principles for Monitoring261
Operational Excellence Pillar261
Security Pillar262
Reliability Pillar263
Performance Efficiency Pillar264
Cost Optimization Pillar266
Sustainability Pillar269
Monitoring Infrastructure and Cost270
Monitoring and Observability Services271
Amazon CloudWatch Logs Insights272
Amazon EventBridge273
AWS CloudTrail274
AWS X-Ray274
Amazon GuardDuty275
Amazon Inspector276
AWS Security Hub277
Cost Tracking and Optimization Services278
AWS Cost Explorer278
AWS Cost and Usage Reports279
AWS Trusted Advisor280
AWS Budgets280
Pricing Models281
Summary283
Exam Essentials284
Review Questions286
Chapter 8Model Security289
Security Design Principles290
Implement a Strong Identity Foundation290
Apply Security at all Layers291
Enable Traceability292
Protect Your Data (At-Rest, In-Use, and In-Transit)293
Automate Security Processes294
Prepare for Security Events295
Securing AWS Services295
Securing Identities with IAM296
Identities296
Access Policies302
Securing Infrastructure and Data305
Network Isolation with VPC305
Private Connectivity306
Data Protection306
Monitoring and Auditing307
Ensuring Compliance307
Summary308
Exam Essentials309
Review Questions311
