Data Science and ML Training Overview
This Applied Data Science and Practical Machine Learning with AWS SageMaker and AutoML training course teaches attendees the latest Machine Learning (ML) techniques. Students learn the fundamentals of ML, including exploratory data analysis, model building, and ML explainability. Participants also learn how to use the latest AutoML tools and techniques, such as auto-sklearn, H2O, Auto-Keras, and AWS Auto Pilot. Finally, attendees learn how to use AWS SageMaker to train, evaluate, and deploy models. This data science ML course also includes advanced topics, including neural networks, deep learning, transfer learning, and fine-tuning.
Location and Pricing
Accelebrate offers instructor-led enterprise training for groups of 3 or more online or at your site. Most Accelebrate classes can be flexibly scheduled for your group, including delivery in half-day segments across a week or set of weeks. To receive a customized proposal and price quote for private corporate training on-site or online, please contact us.
In addition, some courses are available as live, instructor-led training from one of our partners.
Objectives
- Understand the data science life cycle
- Set up a SageMaker environment
- Train and evaluate ML models using SageMaker
- Deploy ML models
- Work with an AWS AutoML or auto-sklearn environment
- Work with ML models using H2O's automated machine learning
- Understand neural networks and deep learning
Prerequisites
- Proficiency in Python programming
- Understanding of data analysis and manipulation techniques
- Familiarity with Python Pandas or Numpy is recommended
- Basic knowledge of Machine Learning concepts, algorithms, and model evaluation
- Familiarity with AWS and some experience with S3, IAM, and EC2 services
Outline
Expand All | Collapse All
Introduction
Data Processing Phases and the Data Science Life Cycle
- Introduction to the data science life cycle
- Data exploration and visualization
- Data cleaning and preprocessing
- Feature engineering
- Model selection and evaluation
- Tuning ML: data, parameters, hyperparameters, and artifacts
- MLI, tuning through data selection/enrichment, analyzing the manifold
- MLI tools and techniques
Working with ML Algorithms on SageMaker
- Introduction to SageMaker
- Setting up a SageMaker environment
- Training and evaluating ML models using SageMaker's built-in algorithms
- Fine-tuning ML models using SageMaker's hyperparameter tuning
Deploying ML Models as Executable Artifacts
- Introduction to deploying ML models as executable artifacts
- Creating and deploying ML models as REST APIs using SageMaker
- Updating and serving ML models using SageMaker's A/B testing and blue/green deployments
AWS AutoML and Auto-sklearn
- Introduction to Canvas and AWS AutoML
- Costs and examples
- AutoML as auto-hyperparameter tuning with auto-sklearn
- Setting up an AWS AutoML or auto-sklearn environment
- Training and evaluating ML models using AWS AutoML or auto-sklearn
- Fine-tuning ML models using AWS AutoML or auto-sklearn's hyperparameter tuning
Automated Machine Learning with H2O
- Fully automated ML (auto parameter tuning and auto feature engineering)
- H2O libraries, driverless AI, etc
- H2O automl vs auto-sklearn (libraries compared/side-by-side0
- Introduction to H2O and its automated machine learning capabilities
- Setting up an H2O environment (mention JRE req’s)
- Training and evaluating ML models using H2O's automated machine learning
- Fine-tuning ML models using H2O's hyperparameter tuning
Neural Networks (NN)
- Neural networks basics and intro
- NN’s as autoML
- Common NN topologies and applications (RNN, CNN, LSTM, etc)
- Thin layer NN, examples, and lab (using TF)
- Deep Learning
- Libraries (Keras vs. TF vs. pytorch)
Conclusion
Training Materials
All Data Science and Machine Learning training students receive courseware covering the topics in the class.
Software Requirements
- A modern web browser and an Internet connection that allows connections by SSH or Remote Desktop (RDP) into AWS virtual machines.
- Windows, Mac, or Linux with at least 8 GB RAM
- A current version of Anaconda for Python 3.x
- Related lab files that Accelebrate will provide
