Implementing a Machine Learning solution with Azure Databricks (DP-3014)

Learn to train, tune and manage machine learning models with Azure Databricks

Description

Azure Databricks is a cloud-scale platform for data analytics and machine learning. Data scientists and machine learning engineers can use Azure Databricks to implement machine learning solutions at scale.

Subjects

• Provision an Azure Databricks workspace.
• Identify core workloads and personas for Azure Databricks.
• Describe key concepts of an Azure Databricks solution.

• Describe key elements of the Apache Spark architecture.
• Create and configure a Spark cluster.
• Describe use cases for Spark.
• Use Spark to process and analyze data stored in files.
• Use Spark to visualize data.

• Prepare data for machine learning
• Train a machine learning model
• Evaluate a machine learning model

• Use MLflow to log parameters, metrics, and other details from experiment runs.
• Use MLflow to manage and deploy trained models.

• Use the Hyperopt library to optimize hyperparameters.
• Distribute hyperparameter tuning across multiple worker nodes.

• Use the AutoML user interface in Azure Databricks
• Use the AutoML API in Azure Databricks

• Train a deep learning model in Azure Databricks
• Distribute deep learning training by using the Horovod library