The Preference Learning Toolbox provides easy access to a collection of preference learning algorithms and the essential steps associated with building predictive models from ordinal data.

System requirements

This project is written in Java but uses JavaFX so the minimum Java version required is 7u13.

Installation

The tool is distributed as a .jar file and no installation is required. If you have the right Java version installed in your computer, the tool can be launched by simply double-clicking on the .jar file.

How to use

The tool implements 5 steps: (1) data loading, (2) data preprocessing, (3) feature selection, (4) model training and (5) result visualisation. Initially, steps 2-5 are locked. Once a dataset is correctly loaded, steps 2-4 are enabled and accessible as tabs.

The interface allows the user to set up the parameters for steps 2-4 in any order and in fact, preprocessing and feature selection are optional. Once all the options are set, the user can run the experiment from the last tab.

A progress bar and training report will be displayed while the preference learning algorithms run, and once ready a results report is presented.

Each of these steps is supported with inline help dialogues accessible on the question mark button at the bottom-left side of the application window.