Machine Learning Engineering on AWS Training & Certification Course

✦ Machine Learning ✦ Machine Learning Engineering ✦ AWS Machine Learning ✦ Amazon SageMaker ✦ MLOps ✦ ML Model Development ✦ ML Model Training ✦ ML Model Deployment ✦ Model Evaluation ✦ Model Monitoring ✦ Feature Engineering ✦ Data Preprocessing ✦ Data Preparation ✦ Data Transformation ✦ Machine Learning Pipelines ✦ SageMaker Pipelines ✦ Model Registry ✦ Real-Time Inference ✦ Batch Inference ✦ Model Optimization ✦ Model Performance Monitoring ✦ Model Drift Detection ✦ Data Drift Detection ✦ ML Automation ✦ CI/CD for Machine Learning ✦ AWS DevOps ✦ AWS Cloud Computing ✦ Amazon S3 ✦ AWS Glue ✦ Amazon Athena ✦ Amazon Redshift ✦ AWS Lambda ✦ Amazon CloudWatch ✦ AWS IAM ✦ AWS VPC ✦ Amazon ECR ✦ AWS Step Functions ✦ Infrastructure as Code ✦ CloudFormation ✦ ML Governance ✦ ML Security ✦ Production Machine Learning ✦ Scalable ML Solutions ✦ AI/ML Workflows ✦ Machine Learning Lifecycle ✦ Cloud-Based Machine Learning ✦ MLOps Automation ✦ Model Operations ✦ Machine Learning Architecture

• ✓ Familiarity with basic machine learning concepts.
• ✓ Working knowledge of Python programming.
• ✓ Familiarity with common Python data science libraries such as NumPy, Pandas and Scikit-learn.
• ✓ Basic understanding of cloud computing concepts.
• ✓ Basic familiarity with AWS services and the AWS Cloud.
• ✓ Basic understanding of data preparation and machine learning workflows.
• ✓ Familiarity with version control systems such as Git is beneficial, although not mandatory.
• ✓ Some exposure to data science or machine learning projects will be helpful.

Module 1: Introduction to Machine Learning on AWS

• Introduction to Machine Learning
• Machine Learning lifecycle
• ML use cases
• Machine learning terminology
• Traditional ML workflows
• Machine Learning on AWS
• Amazon SageMaker AI overview
• AWS services supporting ML workloads
• ML engineering roles and responsibilities
• Responsible Machine Learning

Module 2: Analyzing Machine Learning Challenges

• Identifying business ML problems
• Translating business requirements into ML problems
• Classification problems
• Regression problems
• Clustering problems
• Recommendation use cases
• Supervised learning
• Unsupervised learning
• Training approaches
• Algorithm selection
• Model interpretability
• Balancing performance and business requirements

Module 3: Data Processing for Machine Learning

• Understanding ML data requirements
• Structured and unstructured data
• Data ingestion
• Data storage
• Data processing
• Exploratory Data Analysis
• Data quality
• Handling missing data
• Identifying incorrect data
• Duplicate data
• Data validation
• AWS storage options
• Amazon S3
• Data processing services
• Amazon EMR
• Amazon SageMaker AI data capabilities

Module 4: Data Transformation & Feature Engineering

• Data transformation
• Feature engineering concepts
• Feature selection
• Feature extraction
• Handling missing values
• Handling categorical data
• Data normalization
• Data standardization
• Feature scaling
• Preparing datasets for model training
• Amazon SageMaker Data Wrangler
• SageMaker Processing
• SageMaker Python SDK

Module 5: Choosing a Machine Learning Modeling Approach

• Selecting ML algorithms
• Algorithm characteristics
• Classification algorithms
• Regression algorithms
• Clustering algorithms
• Ensemble methods
• Model selection
• Training requirements
• Model interpretability
• Performance considerations
• Scalability considerations
• Selecting models based on business requirements

Module 6: Training Machine Learning Models

• Machine learning training workflow
• Training datasets
• Validation datasets
• Test datasets
• Training jobs
• Amazon SageMaker AI training
• Training configurations
• Hyperparameters
• Hyperparameter optimization
• Distributed training concepts
• Model artifacts
• Model versioning
• Training optimization

Module 7: Evaluating & Tuning Machine Learning Models

• Model evaluation
• Classification metrics
• Regression metrics
• Precision
• Recall
• F1 score
• Accuracy
• ROC-AUC
• Confusion matrix
• Mean squared error
• Model validation
• Overfitting
• Underfitting
• Bias and variance
• Hyperparameter tuning
• Model optimization
• Model performance analysis

Module 8: Model Deployment Strategies

• ML model deployment
• Real-time inference
• Batch inference
• Asynchronous inference
• Model endpoints
• SageMaker AI endpoints
• Deployment infrastructure
• Scaling ML workloads
• Auto Scaling
• Production deployment considerations
• Deployment strategies
• Model versions
• Model rollback
• Cost considerations

Module 9: Machine Learning Operations — MLOps

• Introduction to MLOps
• ML lifecycle management
• Automation
• Reproducibility
• Model versioning
• Data versioning
• Model pipelines
• Workflow orchestration
• Amazon SageMaker Pipelines
• Continuous Integration
• Continuous Delivery
• ML CI/CD
• Infrastructure automation
• Automated model deployment

Module 10: Automated Deployment & CI/CD

• CI/CD concepts
• CI/CD for ML workflows
• Source control
• Automated testing
• Model validation
• Deployment automation
• Pipeline automation
• AWS development tools
• Infrastructure as Code concepts
• Continuous model delivery
• Automated retraining workflows

Module 11: Securing AWS Machine Learning Resources

• AWS security fundamentals
• Identity and Access Management
• IAM roles
• IAM policies
• Least-privilege access
• Encryption
• Data protection
• Network security
• Secure ML endpoints
• Secure data storage
• Compliance considerations
• Security monitoring
• Protecting ML resources

Module 12: Monitoring Machine Learning Models

• ML model monitoring
• Model performance monitoring
• Data quality monitoring
• Data drift
• Model drift
• Concept drift
• Inference monitoring
• Logging
• Alerting
• Troubleshooting
• Model retraining
• Monitoring infrastructure
• Amazon CloudWatch
• Monitoring ML pipelines

Module 13: Optimizing ML Infrastructure & Costs

• AWS resource optimization
• Compute selection
• Storage optimization
• Cost-aware ML architecture
• Inference optimization
• Scaling strategies
• Monitoring resource utilization
• AWS cost management
• Balancing performance and cost
• Production ML optimization