Prerequisites
- Introduction
- Course Overview
- Why AWS Machine Learning Engineer Associate Certification
- Demo RegisteringTaking an Exam for the First Time What to know
- Demo Machine Learning Engineer Associate Exam Guide What to focus on
Data Preparation for Machine Learning (ML)
- Section Introduction
- Importance of Data Preparation in Machine Learning
- Overview of AWS Data Services for ML
- Data Formats for Machine Learning
- Core AWS Data Sources S3 EFS FSx
- Demo Setting up S3 Buckets for ML Data Lakes
- Streaming Data Sources Kinesis Flink Kafka
- Demo Using Kinesis for Real time Data Ingestion
- Introduction to Data Transformation and Cleaning
- Feature Engineering Techniques
- Encoding Techniques for ML Data
- AWS Glue and Glue DataBrew for Data Transformation
- Demo AWS Glue ETL Jobs for ML Data Preparation
- Working with SageMaker Feature Store
- Demo Creating and Managing Features in SageMaker Feature Store
- Data Quality and Validation
- Pre training Bias Metrics for Different Data Types
- Strategies for Addressing Class Imbalance
- SageMaker Clarify for Bias Detection
- Demo Detecting and Mitigating Bias with SageMaker Clarify
- Data Labeling with SageMaker Ground Truth
- Demo Setting up a Labeling Job with SageMaker Ground Truth
- Data Security Considerations for ML
- Data Encryption Techniques for ML Datasets
- Data Classification Anonymization and Masking
- Compliance Requirements for ML Data PII PHI Data Residency
ML Model Development
- Section Introduction
- Introduction to ML Model Development on AWS
- ML Algorithms and Business Problem Mapping
- AWS AI Services Overview
- Model Interpretability Considerations
- SageMaker Built in Algorithms
- Demo Exploring SageMaker Built in Algorithms
- Foundation Models and SageMaker JumpStart
- Demo Using SageMaker JumpStart for Quick Model Development
- ML Training Process
- Methods to Reduce Model Training Time
- Regularization Techniques in ML
- Hyperparameter Tuning Techniques
- Using SageMaker Script Mode with Frameworks
- Fine Tuning Pre Trained Models
- Demo Fine tuning a Pre trained Model with SageMaker JumpStart
- Ensemble Learning Methods
- Model Size Reduction Techniques
- Demo Model Version Management with SageMaker Model Registry
- Model Evaluation Techniques and Metrics
- Creating Performance Baselines
- Shadow Deployment for Model Evaluation
Deployment and Orchestration of ML Workflows
- Section Introduction
- Introduction to ML Deployment and Orchestration
- ML Deployment Best Practices
- AWS Deployment Services for ML
- Real time vs Batch Inference
- Demo Deploying a Model for Real time Inference
- Demo Setting up Batch Transform Jobs
- SageMaker Endpoint Types Serverless Real time Asynchronous
- Container Options for ML Deployment
- Demo Creating Custom Containers for SageMaker
- Edge Deployment with SageMaker Neo
- On Demand vs Provisioned Resources for ML
- Infrastructure as Code Options for ML Workloads
- Containerization for ML Workloads
- Auto Scaling SageMaker Endpoints
- VPC Configuration for SageMaker
- CICD Principles for ML Workflows
- AWS CodePipeline CodeBuild and CodeDeploy for ML
- Version Control Systems for ML Projects
- Deployment Strategies and Rollback Actions
- Automating Data Ingestion with Orchestration Services
- Demo Setting up SageMaker Pipelines
- Implementing Automated Tests in ML Pipelines
ML Solution Monitoring, Maintenance, and Security
- Section Introduction
- Understanding Drift in ML Models
- Model Performance Monitoring Techniques
- Using SageMaker Model Monitor
- Demo Setting up Model Monitoring with SageMaker
- Detecting Data Drift with SageMaker Clarify
- Demo Implementing Data Drift Detection
- Monitoring Workflows for Anomalies
- AB Testing for Model Performance Monitoring
- Demo Setting up AB Testing with SageMaker
- Key Performance Metrics for ML Infrastructure
- Monitoring Tools for ML Systems
- Demo Using CloudWatch for ML Infrastructure Monitoring
- Using CloudTrail for ML Activity Logging
- ML Specific Instance Types and Performance Considerations
- Cost Analysis and Optimization for ML Workloads
- Demo Using AWS Cost Explorer for ML Cost Tracking
- Rightsizing ML Instances with SageMaker Inference Recommender
- Demo Using SageMaker Inference Recommender
- Optimizing Infrastructure Costs with Spot Instances
- IAM Roles Policies and Groups for ML Services
- SageMaker Security and Compliance Features
- SageMaker Role Manager
- Network Access Controls for ML Resources
- Security Best Practices for ML CICD Pipelines
- Monitoring and Auditing ML Systems