Learning Objectives

• Describe core ML concepts and how they apply within the AWS Cloud.
• Use AWS services to process, transform, and prepare data for ML workloads.
• Choose suitable ML algorithms and modeling approaches based on problem requirements and interpretability needs.
• Build and implement scalable ML pipelines on AWS covering model training, deployment, and orchestration.
• Set up automated CI/CD pipelines for ML workflows.
• Identify appropriate security controls for ML resources on AWS.
• Apply monitoring strategies to deployed ML models, including methods for identifying data drift.

Course Outline

Module Module 0: Course Introduction

[]

Module Module 1: Getting Started with Machine Learning on AWS

[‘Topic A: ML fundamentals’, ‘Topic B: Amazon SageMaker AI overview’, ‘Topic C: Responsible ML practices’]

Module Module 2: Assessing Machine Learning Challenges

[‘Topic A: Identifying and evaluating ML business problems’, ‘Topic B: Approaches to ML model training’, ‘Topic C: Common ML training algorithms’]

Module Module 3: Preparing Data for Machine Learning

[‘Topic A: Data types and preparation techniques’, ‘Topic B: Exploratory data analysis’, ‘Topic C: AWS storage options and how to select among them’]

Module Module 4: Data Transformation and Feature Engineering

[‘Topic A: Addressing incorrect, duplicate, and missing data’, ‘Topic B: Feature engineering fundamentals’, ‘Topic C: Feature selection methods’, ‘Topic D: AWS services for data transformation’, ‘Lab 1: Analyze and Prepare Data with Amazon SageMaker Data Wrangler and Amazon EMR’, ‘Lab 2: Data Processing Using SageMaker Processing and the SageMaker Python SDK’]

Module Module 5: Selecting a Modeling Approach

[‘Topic A: Built-in algorithms available in Amazon SageMaker AI’, ‘Topic B: How to choose a built-in training algorithm’, ‘Topic C: Amazon SageMaker Autopilot’, ‘Topic D: Factors to consider when selecting a model’, ‘Topic E: Cost considerations for ML workloads’]

Module Module 6: Training Machine Learning Models

[‘Topic A: Core model training concepts’, ‘Topic B: Running model training jobs in Amazon SageMaker AI’, ‘Lab 3: Training a model with Amazon SageMaker AI’]

Module Module 7: Evaluating and Tuning Machine Learning Models

[‘Topic A: Measuring and interpreting model performance’, ‘Topic B: Methods for reducing training time’, ‘Topic C: Hyperparameter tuning approaches’, ‘Lab 4: Model Tuning and Hyperparameter Optimization with Amazon SageMaker AI’]

Module Module 8: Model Deployment Strategies

[‘Topic A: Deployment considerations and available target options’, ‘Topic B: Common deployment strategies’, ‘Topic C: Selecting an inference strategy’, ‘Topic D: Container and instance types for model inference’, ‘Lab 5: Shifting Traffic A/B’]

Module Module 9: Securing AWS Machine Learning Resources

[‘Topic A: Access control for ML resources’, ‘Topic B: Network-level access controls for ML environments’, ‘Topic C: Security considerations within CI/CD pipelines’]

Module Module 10: MLOps and Automated Deployment

[‘Topic A: MLOps concepts and principles’, ‘Topic B: Automating testing within CI/CD pipelines’, ‘Topic C: Continuous delivery services on AWS’, ‘Lab 6: Using Amazon SageMaker Pipelines and the Amazon SageMaker Model Registry with Amazon SageMaker Studio’]

Module Module 11: Monitoring Model Performance and Data Quality

[‘Topic A: Identifying drift in ML models’, ‘Topic B: SageMaker Model Monitor capabilities’, ‘Topic C: Monitoring data quality and model quality metrics’, ‘Topic D: Automated remediation and troubleshooting approaches’, ‘Lab 7: Monitoring a Model for Data Drift’]

Module Module 12: Course Wrap-up

[]