View Full Version : New Weka 3.4.14 and 3.6.0 releases

12-18-2008, 10:25 PM
Hi everybody!

New versions of Weka are available for download from the Weka homepage or directly from Sourceforge:

* Weka 3.4.14 - the stable book version available as ZIP, Win32 installer, Win32 installer incl. JRE 1.4.2_13 and Mac OS X application.

* Weka 3.6.0 - stable GUI version created from the head of the development version. The 3.6 code line will receive bug-fixes only from now on. It is available as ZIP, with Win32 installer, Win32 installer incl. JRE 1.5.0_16 and Mac OS X application.

Both versions contain a significant number of bugfixes, it is recommended to upgrade to the new versions.

Weka homepage:

SourceForge site:

Pentaho PDM (Weka) community documentation:

It might take a while before Sourceforge.net has propagated all the files to its mirrors.

What's new in 3.6.0?

Some highlights

* PMML import
* Rotation Forest classifier
* LADTree classifier (multi-class alternating decision trees)
* ScatterSearch (new search algorithm in the attribute selection package)
* Wrapper classifier for the LibLINEAR library

* K-Means now has an option to use the Manhattan distance function in combination with the component-wise median as the centroids.

* The Explorer has some small usability enhancements.
* KnowledgeFlow has major improvements to user-feedback in the form of a new status messaging area.
* Classifier component in the KnowledgeFlow is now multi-threaded and can learn models on multiple cross-validation folds in parallel.
* Improved help for classifiers and clusterers is now available from the command-line interface as well as the GUI's.
* Averaged information retrieval statistics are now available in the GUI's as well as from the command line.
* Weka now has a central logging facility used by all the major user interfaces.
* A new monolithic user manual for Weka, incorporating all the existing individual tutorial pdfs plus info from the WekaDoc wiki.

As usual, for a complete list of changes refer to the changelogs.

The Weka Team