Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
June 1, 1983Journal of the American Statistical Association

Estimating the Error Rate of a Prediction Rule: Improvement on Cross-Validation

View Full Paper
Ask AI
Bookmark
Share

Authors

BEBradley EfronU.S. National Science Foundation

Discussion

Loading...

Member takes

Implication

Statistical evaluation demonstrates bootstrap-based estimators reduce error rate estimation variance in small samples, suggesting substantial improvements over standard cross-validation.

Key Points

  • Clarify the theoretical foundation of predicting classification rule error rates and develop improved error estimators for small-sample datasets.
  • Analyzed the theoretical and mathematical connections linking standard cross-validation to bootstrap estimation for prediction rules.
  • Evaluated alternative bootstrap-derived estimators designed to assess future classification error rates using original sample data.
  • Demonstrated that cross-validation is intimately connected to nonparametric bootstrap estimates of classification error.
  • Identified alternative estimators that offer considerably improved accuracy and lower variance over cross-validation when sample sizes are small.

Cite This Study

Bradley Efron (1983) studied this question.

synapsesocial.com/papers/6a0ccaa09d761985b14a4ad0https://doi.org/10.2307/2288636
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Estimating the Error Rate of a Prediction Rule: Improvement on Cross-Validation1983 · 2,182 citations
  2. 2The Estimation of Prediction Error2004 · 476 citations
  3. 3Improving prediction accuracy by choosing resampling distribution via cross-validation2024
  4. 4Performance and estimation of the true error rate of classification rules built with additional information. An application to a cancer trial2013 · 7 citations
  5. 5The Structure of Cross-Validation Error: Stability, Covariance, and Minimax Limits2025