Classification by probabilistic clustering

This paper describes an approach to classification based on a probabilistic clustering method. Most current classifiers perform classification by modeling class conditional densities directly or by modeling class-dependent discriminant functions. The approach described in this paper uses class-indep...

Full description

Saved in:
Bibliographic Details
Published in2001 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.01CH37221) Vol. 2; pp. 1333 - 1336 vol.2
Main Author Breuel, T.M.
Format Conference Proceeding
LanguageEnglish
Published IEEE 2001
Subjects
Character recognition
Clustering methods
Degradation
Handwriting recognition
Humans
Image coding
Multilayer perceptrons
Optical character recognition software
Robustness
Testing
Online AccessGet full text

Cover

More Information
Summary:This paper describes an approach to classification based on a probabilistic clustering method. Most current classifiers perform classification by modeling class conditional densities directly or by modeling class-dependent discriminant functions. The approach described in this paper uses class-independent multilayer perceptrons (MLP) to estimate the probability that two given feature vectors are in the same class. These probability estimates are used to partition the input into separate classes in a probabilistic clustering. Classification by probabilistic clustering potentially offers greater robustness to different compositions of training and test sets than existing classification methods. Experimental results demonstrating the effectiveness of the method are given for an optical character recognition (OCR) problem. The relationship of the current approach to mixture density estimation, mixture discriminant analysis, and other OCR and handwriting recognition techniques is discussed.
ISBN:0780370414
9780780370418
ISSN:1520-6149
2379-190X
DOI:10.1109/ICASSP.2001.941172