first
prev
toc
next
last
Download Citation
Recommend
Mark
Alert
Link
Print
On the estimation of the expected actual error rate in sample-based multinomial classification
Author:
M. M
hner a
hner a
| Affiliation: | a Central Institute of Cancer Research, National Cancer Registry, Berlin |
DOI:
10.1080/02331888708802057
Publication Frequency:
6 issues per year
Subjects:
Mathematical Statistics;
Statistical Theory & Methods;
Statistics;
Statistics for the Biological Sciences;
Stochastic Models & Processes;
Formats available:
PDF
(English)
Article Requests:
Order Reprints
:
Request Permissions
-
Purchase Article options: (hide options)
-
Sign In
- Online Sample
| purchase type | customer type | online access | payment method | price | |
| Single Article Purchase | Any | 3 days, 1 user, 3 cookies | credit card | US$37.00 |
buy now  add to cart 
|
| Issue Purchase | Any | permanent | credit card | US$539.83 |
buy now add to cart
|
Individual members of the Bernoulli Society are eligible for an online-only discounted personal subscription.
See order form for further information.
View Article:
View Article (PDF)
Abstract
Assuming the two class case, several methods for estimating the expected actual error rate in sample-based multinomial classification are compared with respect to bias and mean absolute error. “Smooth” modifications of the estimators of COCHRAN and HOPKINS (1961) and HILLS (1966) are recommended. Some numerical results illustrate the performance of these estimators particularly in the case of sparsely occupied cells in contingency tables
|
| Keywords: Primary 62H30,62H17; secondary 62E25; Multinomial distributions; error rates; posterior probability estimators; plug-in estimatiors; Bootstrap; leaving-one-out; cross-validation |
| view references (36) |
|
Bookmark with:
|
CiteULike
Del.icio.us
BibSonomy
Connotea
More bookmarks
