Estimating the error rate of a prediction rule: improvement on cross-validation
B Efron - Journal of the American statistical association, 1983 - Taylor & Francis
Journal of the American statistical association, 1983•Taylor & Francis
We construct a prediction rule on the basis of some data, and then wish to estimate the error
rate of this rule in classifying future observations. Cross-validation provides a nearly
unbiased estimate, using only the original data. Cross-validation turns out to be related
closely to the bootstrap estimate of the error rate. This article has two purposes: to
understand better the theoretical basis of the prediction problem, and to investigate some
related estimators, which seem to offer considerably improved estimation in small samples.
rate of this rule in classifying future observations. Cross-validation provides a nearly
unbiased estimate, using only the original data. Cross-validation turns out to be related
closely to the bootstrap estimate of the error rate. This article has two purposes: to
understand better the theoretical basis of the prediction problem, and to investigate some
related estimators, which seem to offer considerably improved estimation in small samples.
Abstract
We construct a prediction rule on the basis of some data, and then wish to estimate the error rate of this rule in classifying future observations. Cross-validation provides a nearly unbiased estimate, using only the original data. Cross-validation turns out to be related closely to the bootstrap estimate of the error rate. This article has two purposes: to understand better the theoretical basis of the prediction problem, and to investigate some related estimators, which seem to offer considerably improved estimation in small samples.
Taylor & Francis Online