Build and operate machine learning solutions with Azure Databricks

0
Microsoft Learn
Free Online Course
English
4-5 hours worth of material
selfpaced

Overview

  • Module 1: Get started with Azure Databricks
  • After completing this module, you will be able to:

    • Understand Azure Databricks
    • Provision Azure Databricks workspaces and clusters
    • Work with notebooks in Azure Databricks
  • Module 2: Work with data in Azure Databricks
  • After completing this module, you will be able to:

    • Understand dataframes
    • Query dataframes
    • Visualize data
  • Module 3: Prepare data for machine learning with Azure Databricks
  • After completing this module, you will be able to:

    • Understand machine learning concepts
    • Perform data cleaning
    • Perform feature engineering
    • Perform data scaling
    • Perform data encoding
  • Module 4: Train a machine learning model with Azure Databricks
  • After completing this module, you will be able to:

    • Understand Spark ML
    • Train and validate a model
    • Use other machine learning frameworks
  • Module 5: Use MLflow to track experiments in Azure Databricks
  • After completing this module, you will be able to:

    • Understand capabilities of MLflow
    • Use MLflow terminology
    • Run experiments
  • Module 6: Manage machine learning models in Azure Databricks
  • After completing this module, you will be able to:

    • Describe considerations for model management
    • Register models
    • Manage model versioning
  • Module 7: Track Azure Databricks experiments in Azure Machine Learning
  • After completing this module, you will be able to:

    • Describe Azure Machine Learning
    • Run Azure Databricks experiments in Azure Machine Learning
    • Log metrics in Azure Machine Learning with MLflow
    • Run Azure Machine Learning pipelines on Azure Databricks compute
  • Module 8: Deploy Azure Databricks models in Azure Machine Learning
  • After completing this module, you will be able to:

    • Describe considerations for model deployment
    • Plan for Azure Machine Learning deployment endpoints
    • Deploy a model to Azure Machine Learning
    • Troubleshoot model deployment
  • Module 9: Tune hyperparameters with Azure Databricks
  • After completing this module, you will be able to:

    • Understand hyperparameter tuning and its role in machine learning.
    • Learn how to use the two open-source tools - automated MLflow and Hyperopt - to automate the process of model selection and hyperparameter tuning.
  • Module 10: Distributed deep learning with Horovod and Azure Databricks
  • After completing this module, you’ll be able to:

    • Understand what Horovod is and how it can help distribute your deep learning models.
    • Use HorovodRunner in Azure Databricks for distributed deep learning.

Syllabus

  • Module 1: Get started with Azure Databricks
    • Introduction
    • Understand Azure Databricks
    • Provision Azure Databricks workspaces and clusters
    • Work with notebooks in Azure Databricks
    • Exercise - Get started with Azure Databricks
    • Knowledge check
    • Summary
  • Module 2: Work with data in Azure Databricks
    • Introduction
    • Understand dataframes
    • Query dataframes
    • Visualize data
    • Exercise - Work with data in Azure Databricks
    • Knowledge check
    • Summary
  • Module 3: Prepare data for machine learning with Azure Databricks
    • Introduction
    • Understand machine learning concepts
    • Perform data cleaning
    • Perform feature engineering
    • Perform data scaling
    • Perform data encoding
    • Exercise - Prepare data for machine learning
    • Knowledge check
    • Summary
  • Module 4: Train a machine learning model with Azure Databricks
    • Introduction
    • Understand Spark ML
    • Train and validate a model
    • Use other machine learning frameworks
    • Exercise - Train a machine learning model
    • Knowledge check
    • Summary
  • Module 5: Use MLflow to track experiments in Azure Databricks
    • Introduction
    • Understand capabilities of MLflow
    • Use MLflow terminology
    • Run experiments
    • Exercise - Use MLflow to track an experiment
    • Knowledge check
    • Summary
  • Module 6: Manage machine learning models in Azure Databricks
    • Introduction
    • Describe considerations for model management
    • Register models
    • Manage model versioning
    • Exercise - Manage models in Azure Databricks
    • Knowledge check
    • Summary
  • Module 7: Track Azure Databricks experiments in Azure Machine Learning
    • Introduction
    • Describe Azure Machine Learning
    • Run Azure Databricks experiments in Azure Machine Learning
    • Log metrics in Azure Machine Learning with MLflow
    • Run Azure Machine Learning pipelines on Azure Databricks compute
    • Exercise - Use Azure Databricks with Azure Machine Learning
    • Knowledge check
    • Summary
  • Module 8: Deploy Azure Databricks models in Azure Machine Learning
    • Introduction
    • Describe considerations for model deployment
    • Plan for Azure Machine Learning deployment endpoints
    • Deploy a model to Azure Machine Learning
    • Troubleshoot model deployment
    • Exercise - Deploy an Azure Databricks model in Azure Machine Learning
    • Knowledge check
    • Summary
  • Module 9: Tune hyperparameters with Azure Databricks
    • Introduction
    • Understand hyperparameter tuning
    • Automated MLflow for model tuning
    • Hyperparameter tuning with Hyperopt
    • Exercise
    • Knowledge check
    • Summary
  • Module 10: Distributed deep learning with Horovod and Azure Databricks
    • Introduction
    • Understand Horovod
    • HorovodRunner for distributed deep learning
    • Exercise
    • Knowledge check
    • Summary