Gaussian mixture pdf in one-class classification: computing and utilizing confidence values

Authors:
J. Ilonen;P. Paalanen;J.-K. Kamarainen;H. Kalviainen
Affiliations:
Lappeenranta University of Technology;Lappeenranta University of Technology;Lappeenranta University of Technology;Lappeenranta University of Technology
Venue:
ICPR '06 Proceedings of the 18th International Conference on Pattern Recognition - Volume 02
Year:
2006

Citing 0
Cited 3

Automatic camera control using unobtrusive vision and audio tracking

Proceedings of Graphics Interface 2010
Novelty Detection with Multivariate Extreme Value Statistics

Journal of Signal Processing Systems
Review: A review of novelty detection

Signal Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this study a confidence measure for probability density functions (pdfs) is presented. The measure can be used in one-class classification to select a pdf threshold for class inclusion. In addition, confidence information can be used to verify correctness of a decision in a multi-class case where for example the Bayesian decision rule reveals which class is the most probable. Additionally, using confidence valueswhich represent in which quantile of the probability mass a pdf value resides ([0, 1]) - is often straightforward compared to using arbitrarily scaled pdf values. As the main contributions, use of confidence information in classification is described and a method for confidence estimation is presented.