KEEL: A software tool to assess evolutionary algorithms for Data Mining problems (regression, classification, clustering, pattern mining and so on)

KEEL-dataset - data set description

Letter Recognition data set

Description
Files and additional references

Description

This section describes main characteristics of the letter data set and its attributes:

General information

Letter Recognition data set
Type	Classification	Origin	Real world
Features	16	(Real / Integer / Nominal)	(0 / 16 / 0)
Instances	20000	Classes	26
Missing values?			No

Attribute description

Attribute	Domain	Attribute	Domain
X-box	[0, 15]	Y2bar	[0, 15]
Y-box	[0, 15]	Xybar	[0, 15]
Width	[0, 15]	X2ybr	[0, 15]
High	[0, 15]	Xy2br	[0, 15]
Onpix	[0, 15]	X-ege	[0, 15]
X-bar	[0, 15]	Xegvy	[0, 15]
Y-bar	[0, 15]	Y-ege	[0, 15]
X2bar	[0, 15]	Yegvx	[0, 15]
Class	{A, ..., Z}

Additional information

The objective is to identify each of a large number of black-and-white rectangular pixel displays as one of the 26 capital letters in the English alphabet. The character images were based on 20 different fonts and each letter within these 20 fonts was randomly distorted to produce a file of 20,000 unique stimuli. Each stimulus was converted into 16 primitive numerical attributes (statistical moments and edge counts) which were then scaled to fit into a range of integer values from 0 through 15.

Files and additional references

In this section you can download some files related to the letter data set:

The complete data set already formatted in KEEL format can be downloaded from here .
A copy of the data set already partitioned by means of a 10-folds cross validation procedure can be downloaded from here .
A copy of the data set already partitioned by means of a 5-folds cross validation procedure can be downloaded from here .
The header file associated to this data set can be downloaded from here .
This is not a native data set from the KEEL project. It has been obtained from the UCI Machine Learning Repository. The original page where the data set can be found is: http://archive.ics.uci.edu/ml/datasets/Letter+Recognition.