Feature Selection Based on Class Separation in Handwritten Numeral Recognition Using Neural Network


The Transactions of the Korea Information Processing Society (1994 ~ 2000), Vol. 6, No. 2, pp. 543-551, Feb. 1999
10.3745/KIPSTE.1999.6.2.543,   PDF Download:

Abstract

The primary purposes in this paper are to analyze the class separtion of features in handwritten numeral recognition and to make use of the results in feature selection. Using the Parzen window technique, we compute the class distributions and define the class separation to be the overlapping distance of two class distributions. The dimension of a feature vector is reduced by removing the void or redundant feature cells based on the class separation information. The experiments have been performed on the CENPARMI handwritten numeral database, and partial classification and full classification have been tested. The results show that the class separation is very effective for the feature selection in the 10-class handwritten numeral recognition problem since we could reduce the dimension of the original 256-dimensional feature vector by 22%.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
L. J. Seon, "Feature Selection Based on Class Separation in Handwritten Numeral Recognition Using Neural Network," The Transactions of the Korea Information Processing Society (1994 ~ 2000), vol. 6, no. 2, pp. 543-551, 1999. DOI: 10.3745/KIPSTE.1999.6.2.543.

[ACM Style]
Lee Jin Seon. 1999. Feature Selection Based on Class Separation in Handwritten Numeral Recognition Using Neural Network. The Transactions of the Korea Information Processing Society (1994 ~ 2000), 6, 2, (1999), 543-551. DOI: 10.3745/KIPSTE.1999.6.2.543.