DocumentCode
918886
Title
Strong consistency of an estimate of the asymptotic error probability of the nearest neighbor rule (Corresp.)
Author
Wolverton, Charles T.
Volume
19
Issue
1
fYear
1973
fDate
1/1/1973 12:00:00 AM
Firstpage
119
Lastpage
120
Abstract
Let 
be the asymptotic error probability of the nearest neighbor rule (NNR). A particular estimate of from 
-classified samples, known to converge to in probability as increases to infinity, is shown to actually converge with probability 1 to whenever the NNR itself converges with probability 1.
be the asymptotic error probability of the nearest neighbor rule (NNR). A particular estimate of from -classified samples, known to converge to in probability as increases to infinity, is shown to actually converge with probability 1 to whenever the NNR itself converges with probability 1.Keywords
Estimation; Pattern classification; Error probability; H infinity control; Lagrangian functions; Layout; Maximum likelihood estimation; Nearest neighbor searches; Nonlinear equations; Nonlinear optics; Optical sensors; Relaxation methods;
fLanguage
English
Journal_Title
Information Theory, IEEE Transactions on
Publisher
ieee
ISSN
0018-9448
Type
jour
DOI
10.1109/TIT.1973.1054947
Filename
1054947
Link To Document