A comparative study of feature ranking methods in recognition of handwritten numerals

N/ACitations
Citations of this article
9Readers
Mendeley users who have this article in their library.
Get full text

Abstract

Feature selection is an important task during classification of any pattern. In this paper, we compute and compare the strengths of five most widely used feature ranking techniques in identifying the optimal subset of features for best classification results. The feature ranking measurements that are used here are information gain (IG), gain ratio (GR), correlation, symmetrical uncertainty (SU), and chi-square (CS). For evaluation purpose, recognition of handwritten numeral samples from five popular Indic scripts—Bangla, Hindi, English, Telugu, and Arabic—are used. These ranking methods are applied over quadtree-based longest run feature set. Experimental results are drawn and compared using support vector machine (SVM)-based classifier.

Cite

CITATION STYLE

APA

Roy, A., Das, N., Saha, A., Sarkar, R., Basu, S., Kundu, M., & Nasipuri, M. (2015). A comparative study of feature ranking methods in recognition of handwritten numerals. In Advances in Intelligent Systems and Computing (Vol. 324, pp. 473–479). Springer Verlag. https://doi.org/10.1007/978-81-322-2126-5_52

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free