Practical Data Science with Amazon SageMaker

Module 1: Introduction to Data Science on AWS

• Data science fundamentals
• Machine learning lifecycle
• Data science workflows
• AWS machine learning ecosystem
• Amazon SageMaker overview
• SageMaker Studio
• Machine learning use cases
• AWS data science architecture

Module 2: Data Preparation and Exploration

• Data collection
• Data ingestion
• Dataset exploration
• Data cleaning
• Missing data
• Outlier detection
• Data transformation
• Data normalization
• Exploratory Data Analysis
• Amazon S3 for data storage

Module 3: Feature Engineering

• Feature engineering fundamentals
• Selecting relevant features
• Feature transformation
• Encoding categorical data
• Feature scaling
• Handling missing values
• Feature selection
• Creating training datasets
• Preparing data for SageMaker

Module 4: Machine Learning Model Development

• Supervised learning
• Unsupervised learning
• Classification
• Regression
• Clustering
• Model selection
• Training and validation datasets
• SageMaker built-in algorithms
• Model development workflow

Module 5: Model Training with Amazon SageMaker

• SageMaker training jobs
• Training instances
• Training datasets
• Model artifacts
• Hyperparameters
• SageMaker built-in algorithms
• Custom training workflows
• Training performance
• Distributed training concepts

Module 6: Model Evaluation and Tuning

• Model evaluation
• Performance metrics
• Accuracy
• Precision and recall
• F1 score
• Confusion matrix
• Regression metrics
• Cross-validation
• Overfitting and underfitting
• Hyperparameter tuning
• Model optimization

Module 7: Machine Learning with SageMaker Algorithms

• SageMaker built-in algorithms
• XGBoost
• Linear Learner
• K-Means
• Principal Component Analysis
• Classification models
• Regression models
• Algorithm selection
• Model performance comparison

Module 8: Model Deployment and Inference

• Model packaging
• SageMaker model deployment
• Real-time inference
• Batch transform
• Inference endpoints
• Endpoint configuration
• Auto Scaling
• Serverless inference
• Prediction workflows
• Endpoint security

Module 9: Machine Learning Pipelines

• ML workflow automation
• SageMaker Pipelines
• Data processing steps
• Training steps
• Evaluation steps
• Model registration
• Automated workflows
• Reproducibility
• Pipeline monitoring

Module 10: Model Monitoring

• Production model monitoring
• Data quality
• Model quality
• Data drift
• Model drift
• SageMaker Model Monitor
• CloudWatch integration
• Monitoring metrics
• Alerts
• Model maintenance

Module 11: Security and Governance

• AWS IAM
• SageMaker security
• Roles and permissions
• Data encryption
• Amazon S3 security
• Network security
• VPC integration
• Data privacy
• Model governance
• Responsible machine learning

Module 12: Hands-On Data Science Project

• Define a machine learning problem
• Prepare and explore a dataset
• Engineer features
• Train a machine learning model
• Evaluate model performance
• Tune hyperparameters
• Deploy the model
• Generate predictions
• Monitor the deployed model
• Analyze model performance
• Document the complete ML workflow

• Data Scientists
• Machine Learning Engineers
• Data Engineers
• ML Engineers
• MLOps Engineers
• Software Developers working with machine learning
• AI Engineers
• Cloud Engineers
• DevOps Professionals supporting ML workloads
• Business Analysts transitioning into data science
• Data Science Professionals moving to AWS
• Professionals preparing for AWS machine learning certifications
• Professionals responsible for developing and deploying machine learning solutions on AWS

• Understand the complete data science and machine learning lifecycle.
• Use Amazon SageMaker to develop and manage machine learning models.
• Prepare, clean, transform, and analyze datasets.
• Apply effective feature engineering techniques.
• Select appropriate machine learning algorithms.
• Train models using SageMaker training jobs.
• Evaluate models using appropriate performance metrics.
• Perform hyperparameter tuning and model optimization.
• Deploy models using real-time and batch inference.
• Configure and manage SageMaker inference endpoints.
• Build automated workflows using SageMaker Pipelines.
• Monitor deployed models for data quality and model drift.
• Apply AWS IAM and security practices to ML workloads.
• Develop reproducible and scalable machine learning workflows.
• Gain hands-on experience building an end-to-end machine learning project on AWS.

• Basic knowledge of statistics and probability.
• Understanding of fundamental machine learning concepts.
• Familiarity with Python programming.
• Basic understanding of data preparation and analysis.
• Familiarity with common machine learning algorithms.
• Basic knowledge of AWS Cloud services.
• Understanding of Amazon S3 is beneficial.
• Experience with data analysis tools such as Pandas and NumPy is recommended.
• Basic knowledge of Jupyter notebooks is helpful.

Course: Practical Data Science with Amazon SageMaker
Delivery: Instructor-led, hands-on training
Hands-On: Yes
Technology: Amazon Web Services
Primary Platform: Amazon SageMaker
Focus: Data science, machine learning, model development, deployment, and monitoring