DocumentCode
1800495
Title
Estimation of error rates using smoothed estimators
Author
Pawlak, M. ; Liao, X.
Author_Institution
Dept. of Electr. Eng., Manitoba Univ., Winnipeg, Man., Canada
fYear
1988
fDate
14-17 Nov 1988
Firstpage
954
Abstract
A continuity-corrected version of the commonly used counting estimates is examined. The estimate to be studied has a reduced bias as well as variance. Moreover, it does not require intensive resampling techniques (as in the case of bootstrapping) and is therefore simple to compute. Monte Carlo simulations were performed to get insight into the problem of choosing of a smoothing level as well as the selection of a smoothing function
Keywords
Monte Carlo methods; error analysis; estimation theory; pattern recognition; picture processing; Monte Carlo simulations; bootstrapping; counting estimates; error rates estimation; pattern recognition; resampling techniques; smoothing function; Error analysis; Estimation error; Gaussian distribution; Mean square error methods; Monte Carlo methods; Pattern analysis; Pattern recognition; Performance analysis; Random variables; Smoothing methods;
fLanguage
English
Publisher
ieee
Conference_Titel
Pattern Recognition, 1988., 9th International Conference on
Conference_Location
Rome
Print_ISBN
0-8186-0878-1
Type
conf
DOI
10.1109/ICPR.1988.28412
Filename
28412
Link To Document