Led dataset

This is the well known LED example from the CART book. It is a 10-class classification problem. There are 7 binary inputs, meant to simulate the seven filaments in a LED digit display. Each filament can flip its state with probability 0.1. A second prototask was defined on this dataset which includes 17 additional, binary inputs which act simply as noise on the input. The data was generated using David Aha's program from the UCI repository of machine learning databases. Other information is available in the original documentation that was supplied with the program.

Dataset profile:

Origin: artificial

Usage: historical

Number of attributes: 8

Number of cases: 6,600

Number of prototasks: 1

Number of methods run on this dataset: 0

Download led.tar.gz

Contributed by: Michael Revow using David Aha's program available from the UCI repository of machine learning databases.


Breiman L, Friedman JH, Olshen, RA and Stone CJ. Classification and Regression Trees. Wadsworth International Group: Belmont, California.

Last Updated 6 November 1996
Comments and questions to: delve@cs.toronto.edu
 Copyright