Open Access
Translator Disclaimer
March 2011 Construction and evaluation of classifiers for forensic document analysis
Christopher P. Saunders, Linda J. Davis, Andrea C. Lamas, John J. Miller, Donald T. Gantz
Ann. Appl. Stat. 5(1): 381-399 (March 2011). DOI: 10.1214/10-AOAS379

Abstract

In this study we illustrate a statistical approach to questioned document examination. Specifically, we consider the construction of three classifiers that predict the writer of a sample document based on categorical data. To evaluate these classifiers, we use a data set with a large number of writers and a small number of writing samples per writer. Since the resulting classifiers were found to have near perfect accuracy using leave-one-out cross-validation, we propose a novel Bayesian-based cross-validation method for evaluating the classifiers.

Citation

Download Citation

Christopher P. Saunders. Linda J. Davis. Andrea C. Lamas. John J. Miller. Donald T. Gantz. "Construction and evaluation of classifiers for forensic document analysis." Ann. Appl. Stat. 5 (1) 381 - 399, March 2011. https://doi.org/10.1214/10-AOAS379

Information

Published: March 2011
First available in Project Euclid: 21 March 2011

zbMATH: 1220.62171
MathSciNet: MR2810402
Digital Object Identifier: 10.1214/10-AOAS379

Keywords: Bayesian statistics , ‎classification‎ , cross-validation , handwriting identification

Rights: Copyright © 2011 Institute of Mathematical Statistics

JOURNAL ARTICLE
19 PAGES


SHARE
Vol.5 • No. 1 • March 2011
Back to Top