jaDTi - Decision Trees: a Java implementation

Overview

This site describes jaDTi, a Java implementation of decision trees. It's been designed to be easy to use and general purpose. It is licenced under GPL, thus its source code is freely available.

This library is reasonably efficient, meaning that the complexity of the implementation of the algorithms involved is that given by the theory. However, when a choice must be made between code readability and efficiency, readability has been chosen. It is thus ideal in research (because algorithms can easily be modified) and as an academic tool (students can quickly get interesting results). Its general philosophy can be compared to that of the jahmm library (an implementation of Hidden Markov Models in Java).

This page is not an introduction to decision tree. The reader should have a (basic) idea of what a decision tree is.

Example

This example shows how to build a decision tree using the elements of a database and how to use it to guess an attribute value.

Documentation

The Javadoc of the latest version of these packages is available online, as a tar.gz file or as a zip file.

Download

Download the current release (sources tar.gz or zip, jar) or pick one in the following list (the number given in the filename matches the version):

Version Javadoc Release date Comment
0.5.0 (sources tar.gz or zip, jar) Online, tar.gz, zip October 2004 First public release.
0.5.1 (sources tar.gz or zip, jar) Online, tar.gz, zip November 2004 Various improvements.
0.6.0 (sources tar.gz or zip, jar) Online, tar.gz, zip November 2004 Handles unknown values.
0.6.1 (sources tar.gz or zip, jar) Online, tar.gz, zip November 2004 Bugfixes.

The changelog summarizes the changes since the first version. Please contact Jean-Marc François (the author) at the address francois-jaDTi@run.montefiore.ulg.ac.be if you have any comment or question.