In association with heise online

03 November 2010, 17:57

Apache Mahout 0.4 adds new algorithms

  • Twitter
  • Facebook
  • submit to slashdot
  • StumbleUpon
  • submit to reddit

Apache Mahout Logo The Apache Mahout developers have released version 0.4 of their library of scalable machine-learning algorithms now with a variety of new algorithms, including Spectral Clustering, MinHash Clustering and a new Naive Bayes classifier, as well as new options for evaluating clustering results. Many existing code segments were updated to improve their performance or extend their functionality.

Apache Mahout offers stable implementations of machine-learning algorithms which a focus on processing large amounts of data, even across multiple computers. Typical applications include web-store log analyses for generating "You may also be interested in" lists, grouping texts according to their content, classifying documents in pre-set categories and determining contexts in large data sets. Mahout was initially a sub-project of the Lucene search engine and became a top-level Apache project last spring.

More details about the release can be found in the official release announcement. At the time of this posting, the release notes linked to on the project's home page and in the release announcement still show version 0.3. Apache Mahout 0.4 is available to download from one of the project's mirrors and is licensed under the Apache License, version 2.0.

See also:

(crve)

Print Version | Send by email | Permalink: http://h-online.com/-1130068
 


  • July's Community Calendar





The H Open

The H Security

The H Developer

The H Internet Toolkit