MADlib®: Big Data Machine Learning in SQL for Data Scientists

  • Open Source, commercially usable BSD license
  • Supports Postgres, Pivotal Greenplum Database, and Pivotal HAWQ®
  • Powerful analytics for Big Data

Read More

Latest News

MADlib v1.7.1 Release Announcement

MADlib v1.7.1 is released and available for download.

New features include:

  1. Major performance improvements in Random Forest: forest_train() takes about 30% less time and variable importance computation takes 90% less time.
  2. Added continuous variables for Naive Bayes using Gaussian smoothing to estimate feature probabilities.
  3. Added support for PostgreSQL 9.4.

For a more detailed list of changes see the MADlib v1.7.1 Release Notes.

Access the binaries on the MADlib Download Page. As always the MADlib user forum is open for questions.

MADlib v1.7 Release Announcement

MADlib v1.7 is released and available for download.

New features include:

  1. A new GLM (generalized linear models) module that allows various regression and classification methods beyond the basic linear and logistic regression.
  2. Revamped Decision Tree and Random Forest modules that provide better features, easier-to-use interfaces and better performance.
  3. Enhanced PMML support for exporting GLM output and tree/forest models.

For a more detailed list of changes see the MADlib v1.7 Release Notes.

Older News