Features selection in character recognition with random forest classifier

Authors:
Wladyslaw Homenda;Wojciech Lesinski
Affiliations:
Faculty of Mathematics and Information Science, Warsaw University of Technology, Warsaw, Poland and Faculty of Mathematics and Computer Science, University of Bialystok, Bialystok, Poland;Faculty of Mathematics and Computer Science, University of Bialystok, Bialystok, Poland
Venue:
ICCCI'11 Proceedings of the Third international conference on Computational collective intelligence: technologies and applications - Volume Part I
Year:
2011

Citing 4
Cited 0

Bagging predictors

Machine Learning
Random Forests

Machine Learning
Induction of Decision Trees

Machine Learning
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)

Quantified Score

Hi-index	0.00

Visualization

Abstract

Proper image recognition depends on many factors. Features' selection and classifiers are most important ones. In this paper we discuss a number of features and several classifiers. The study is focused on how features' selection affects classifier efficiency with special attention given to random forests. Different construction methods of decision trees are considered. Others classifiers (k nearest neighbors, decision trees and classifier with Mahalanobis distance) were used for efficiency comparison. Lower case letters from Latin alphabet are used in empirical tests of recognition efficiency.