Classification of style-constrained pattern-fields

Prateek Sarkar, George Nagy

Abstract

In many classification tasks, entire fields of patterns such as images of postal address or ZIP-codes originate from the same, but unknown, source. The class-conditional feature distributions depend on the source of the patterns. Several sources may share the same distribution, or style. The style-conditional distributions are estimated from the training set. The optimal field-classifier computes the class-conditional field-feature-probabilities as the sum of class-and-style-conditional field-feature-probabilities, weighted by the prior probabilities of the styles. We compare the decision regions and error rates of style-weighted classification with both conventional singlet and top-style classification in a minimal family of examples, and discuss some related practical considerations.

Download paper

PostScript (300K)

Bibtex entry

@inproceedings{sarkar:icpr2000
, author = "P. Sarkar and G. Nagy"
, title = "Classification of style-constrained pattern-fields"
, booktitle = "Proceedings of the fifteenth ICPR"
, publisher = "IEEE Computer Society Press"
, address = "Barcelona"
, pages = "859-862"
, year = "2000"
}
Prateek Sarkar
Last modified: Wed Mar 7 16:56:13 PST 2001