Publication - Improving the Classification Accuracy of the Scanning n-tuple Method.
RESEARCH

Improving the Classification Accuracy of the Scanning n-tuple Method.

Research Area:  
Other topics in Computer Science
    
Type:  
In Proceedings

 

Year: 2000
Authors: George Tambouratzis
Editor: Sanfeliu, A., Villanueva, J.J., Vanrell, M., Alquezar, R., Jain, A.K., & Kittler, J
Volume: 2
Book title: Proceedings of the 15th International Conference on Pattern Recognition
Pages: 1050-1053
Address: Barcelona, Spain
Organization: IAPR
Date: 3-8 September
ISBN: 0-7695-0750-6
DOI: 10.1109/ICPR.2000.906254
Abstract:
In this article, the application of the scanning n-tuple technique to classification tasks is studied. The performance of this technique is examined in a handwritten character recognition task where the accuracy is initially low. This task is employed as a case study for designing a general-purpose algorithm that improves the scanning n-tuple performance in hard classification tasks, by focusing on the characteristics of the pattern space. Experimental results indicate that the use of the algorithm results in a substantial improvement of the scanning n-tuple classification performance in comparison to previous results. This improvement is shown to be equivalent to that achieved by employing structural knowledge regarding the specific pattern space
[Bibtex]