mfdigits: Multiple feature digits data set.

Description Usage Format Source References

Description

This dataset consists of features derived from handwritten digits (0–9) taken from Dutch utility maps. We have included only the 210 profile correlation features.

Usage

1

Format

A list with entries $x (a 2000x210 matrix with each row corresponding to an individual handwritten digit) and $c (a vector of labels)

Source

UCI Machine Learning Repository.

References

Dheeru, D. and E. Karra Taniskidou (2017). UCI machine learning repository. https://archive.ics.uci.edu/ml

M. van Breukelen, R.P.W. Duin, D.M.J. Tax, and J.E. den Hartog, Handwritten digit recognition by combined classifiers, Kybernetika, vol. 34, no. 4, 1998, 381-386.


DavidHofmeyr/spuds documentation built on May 20, 2019, 9:40 a.m.