Practical Automated Machine Learning on Azure by Mukunthu, Deepak -- Read -- Imperial Library of Trantor

Index

Foreword Preface

Conventions Used in This Book Using Code Examples O’Reilly Online Learning How to Contact Us Acknowledgments

I. Automated Machine Learning 1. Machine Learning: Overview and Best Practices

Machine Learning: A Quick Refresher

Model Parameters Hyperparameters

Best Practices for Machine Learning Projects

Understand the Decision Process Establish Performance Metrics Focus on Transparency to Gain Trust Embrace Experimentation Don’t Operate in a Silo

An Iterative and Time-Consuming Process

Feature Engineering Algorithm Selection Hyperparameter Tuning The End-to-End Process

Growing Demand Conclusion

2. How Automated Machine Learning Works

What Is Automated Machine Learning?

Understanding Data Detecting Tasks Choosing Evaluation Metrics Feature Engineering

Detect issues with input data and automatically flag them Drop columns that are not useful as features Generate features Select the most impactful features

Selecting a Model

Brute-force approaches Smarter approaches

Monitoring and Retraining Bringing It All Together

Automated ML

How Automated ML Works Preserving Privacy Enabling Transparency Guardrails End-to-End Model Life-Cycle Management

Conclusion

II. Automated ML on Azure 3. Getting Started with Microsoft Azure Machine Learning and Automated ML

The Machine Learning Process

Collaboration and Monitoring Deployment

Setting Up an Azure Machine Learning Workspace for Automated ML

Azure Notebooks Notebook VM

Conclusion

4. Feature Engineering and Automated Machine Learning

Data Preprocessing Methods Available in Automated ML Auto-Featurization for Automated ML

Auto-Featurization for Classification and Regression Auto-Featurization for Time-Series Forecasting

Conclusion

5. Deploying Automated Machine Learning Models

Deploying Models

Registering the Model Creating the Container Image Deploying the Model for Testing Testing a Deployed Model Deploying to AKS

Swagger Documentation for the Web Service Debugging a Deployment

Web Service Deployment Fails

Conclusion

6. Classification and Regression

What Is Classification and Regression?

Classification and Regression Algorithms Using Automated ML for Classification and Regression

Setting up the Azure Machine Learning workspace Data preparation Using automated ML to train the model Selecting and testing the best model from the experiment run

Conclusion

III. How Enterprises Are Using Automated Machine Learning 7. Model Interpretability and Transparency with Automated ML

Model Interpretability

Model Interpretability with Azure Machine Learning

Explainers Regression model trained using sklearn Classification model trained using automated ML

Model Transparency

Understanding the Automated ML Model Pipelines Guardrails

Conclusion

8. Automated ML for Developers

Azure Databricks and Apache Spark ML.NET SQL Server Conclusion

9. Automated ML for Everyone

Azure Portal UI Power BI

Preparing the Data Automated ML Training Understanding the Best Model Understanding the Automated ML Training Process Model Deployment and Inferencing

Enabling Collaboration

Azure Machine Learning to Power BI Power BI Automated ML to Azure Machine Learning

Conclusion

Index

← Prev
Back
Next →

← Prev
Back
Next →