Title :
Bayesian classification and classification performance for independent distributions (Corresp.)
Author :
Kokolakis, George E.
fDate :
7/1/1981 12:00:00 AM
Abstract :
Some general conditions are derived for the mean recognition accuracy of a Bayesian classifier to approach unity as the feature dimensionality increases to infinity.
Keywords :
Bayes procedures; Pattern classification; Bayesian methods; Convergence; H infinity control; Mathematics; Notice of Violation; Performance gain; Recursive estimation; Statistics; Stochastic processes; Unsupervised learning;
Journal_Title :
Information Theory, IEEE Transactions on
DOI :
10.1109/TIT.1981.1056369
