Information criterion for discriminating two events with different occurrence probability

Authors :: S. Goto
M. Nakamura
Source :: Proceedings IEEE International Symposium on Information Theory.
Publication Year :: 2003
Publisher :: IEEE, 2003.
Abstract: In this paper, an information criterion for discriminating two events is proposed. The proposed information criterion is defined by considering the occurrence probability of events. Properties of the criterion are investigated and the criterion is applied to the evaluation and factor selection of the previously proposed discriminant analysis.

Subjects :: business.industry
Occurrence probability
Pattern recognition
Probability density function
Information theory
Linear discriminant analysis
Bayesian information criterion
Statistics
Probability distribution
Statistical analysis
Artificial intelligence
Factor selection
business
Mathematics

Tools