Predictive inference with the jackknife+
From MaRDI portal
DOI10.1214/20-AOS1965zbMATH Open1461.62050OpenAlexW3126856044MaRDI QIDQ100525FDOQ100525
Abstract: This paper introduces the jackknife+, which is a novel method for constructing predictive confidence intervals. Whereas the jackknife outputs an interval centered at the predicted response of a test point, with the width of the interval determined by the quantiles of leave-one-out residuals, the jackknife+ also uses the leave-one-out predictions at the test point to account for the variability in the fitted regression function. Assuming exchangeable training samples, we prove that this crucial modification permits rigorous coverage guarantees regardless of the distribution of the data points, for any algorithm that treats the training points symmetrically. Such guarantees are not possible for the original jackknife and we demonstrate examples where the coverage rate may actually vanish. Our theoretical and empirical analysis reveals that the jackknife and the jackknife+ intervals achieve nearly exact coverage and have similar lengths whenever the fitting algorithm obeys some form of stability. Further, we extend the jackknife+ to K-fold cross validation and similarly establish rigorous coverage properties. Our methods are related to cross-conformal prediction proposed by Vovk [2015] and we discuss connections.
Recommendations
Cites work
- scientific article; zbMATH DE number 3483405 (Why is no real title available?)
- scientific article; zbMATH DE number 2168212 (Why is no real title available?)
- scientific article; zbMATH DE number 3433376 (Why is no real title available?)
- scientific article; zbMATH DE number 3053501 (Why is no real title available?)
- 10.1162/153244302760200704
- A data-driven software tool for enabling cooperative information sharing among police departments
- Bootstrap Prediction Intervals for Regression
- Bootstrap methods: another look at the jackknife
- Cross-conformal predictors
- Distribution-free inequalities for the deleted and holdout error estimates
- Distribution-free predictive inference for regression
- Fast exact conformalization of the Lasso using piecewise linear homotopy
- NOTES ON BIAS IN ESTIMATION
- On conditional moments of high-dimensional random vectors given lower-dimensional projections
- Predictive Intervals Based on Reuse of the Sample
- Scikit-learn: machine learning in Python
- The Predictive Sample Reuse Method with Applications
Cited in
(26)- Root-finding approaches for computing conformal prediction set
- A conformal test of linear models via permutation-augmented regressions
- Conformal Prediction: A Gentle Introduction
- Adaptive novelty detection with false discovery rate guarantee
- Conformal Sensitivity Analysis for Individual Treatment Effects
- Optimal Subsampling via Predictive Inference
- Conformal prediction: a unified review of theory and new challenges
- Nested conformal prediction sets for classification with applications to probation data
- A confidence machine for sparse high-order interaction model
- Multi-split conformal prediction via Cauchy aggregation
- JANOS: An Integrated Predictive and Prescriptive Modeling Framework
- Regression trees for fast and adaptive prediction intervals
- Distribution-Free Prediction Sets for Two-Layer Hierarchical Models
- Distribution-free predictive inference for regression
- Post-selection inference via algorithmic stability
- Conformal prediction beyond exchangeability
- Testing for outliers with conformal p-values
- Distribution-free conditional median inference
- Training-conditional coverage for distribution-free predictive inference
- Incorporating relative error criterion to conformal prediction for positive data
- conformalbayes
- Homeostasis phenomenon in conformal prediction and predictive distribution functions
- Conditional predictive inference for stable algorithms
- On the power of conditional independence testing under model-X
- Fast and Exact Leave-One-Out Analysis of Large-Margin Classifiers
- Multi split conformal prediction
This page was built for publication: Predictive inference with the jackknife+
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q100525)
