DePaul University DePaul CTI Homepage

Data Mining with WEKA

This guide/tutorial uses a detailed example to illustrate some of the basic data preprocessing and mining operations that can be performed using WEKA. It is based on WEKA version 3.4.1. Some of the interface elements and modules may have changed in the most current version of WEKA. You can download the most current version of WEKA from the WEKA Web site. The current version includes a few additional features in the GUI and has a more organized packaging structure for the Java components. You should pay attention to these differences as you go through the tutorial. The differences in packaging structure are particularly important when you are running WEKA from the commandline.

The following topics are available (it is best to follow these in order, if you are not familiar with WEKA 3.4 already).

Topics

  1. Data Preprocessing in WEKA
  2. Association Rule Mining with WEKA
  3. Classification via Decision Trees in WEKA
  4. K-Means Clustering in WEKA

Some additional documents related to WEKA

The official WEKA Web site, including additional resources and sample data sets.
Attribute-Relation File Format (ARFF)
A presentation demonstrating graphical user interfaces in Weka.
An Example of a Simple Message Classifier in Java, Using WEKA Modules
An Introduction to using WEKA from the command line (by Alex Seewald)

Return to Main Page


Copyright © 2005-2006, Bamshad Mobasher, School of CTI, DePaul University.