Introduction to Databricks Machine Learning

This article is an introduction to Databricks Machine Learning. It describes the benefits of using Databricks for common ML tasks and provides links to notebooks, tutorials, and user guides to help you get started.

What is Databricks Machine Learning?

Databricks Machine Learning provides an integrated machine learning environment that helps you simplify and standardize your ML development processes. With Databricks Machine Learning, you can:

• Train models either manually or with AutoML.

• Track training parameters and model performance using experiments with MLflow tracking.

• Share, manage, and serve models using Model Registry and Databricks Model Serving.

• Use the Databricks Feature Store to develop and share features, track upstream and downstream feature lineage, and serve feature values online.

Databricks Machine Learning also includes all of the capabilities of the Databricks workspace, including:

• Data exploration, management, and governance.

• Cluster creation and management.

• Collaborative notebooks.

• Library and environment management.

• Code management with Databricks Repos.

• Automation support including Delta Live Tables, Databricks Jobs, and APIs.

For machine learning applications, Databricks recommends using a cluster running Introduction to Databricks Runtime for Machine Learning.

Use Databricks for model development and deployment

The diagram shows how the capabilities of Databricks map to the steps of the model development and deployment process.

Next steps

To get started, see:

• Databricks Machine Learning home page

• Databricks Machine Learning in-product quickstart

• Tutorials: Get started with ML

To learn about key Databricks Machine Learning features, see:

• What is AutoML?

• Databricks Feature Store

• Model serving with Databricks

For a recommended MLOps workflow on Databricks Machine Learning, see:

• MLOps workflow on Databricks