Distribution-Free Prediction Sets for Two-Layer Hierarchical Models
From MaRDI portal
Abstract: We consider the problem of constructing distribution-free prediction sets for data from two-layer hierarchical distributions. For iid data, prediction sets can be constructed using the method of conformal prediction. The validity of conformal prediction hinges on the exchangeability of the data, which does not hold when groups of observations come from distinct distributions, such as multiple observations on each patient in a medical database. We extend conformal methods to this hierarchical setting. We develop CDF pooling, single subsampling, and repeated subsampling approaches to construct prediction sets in unsupervised and supervised settings. We compare these approaches in terms of coverage and average set size. If asymptotic coverage is acceptable, we recommend CDF pooling for its balance between empirical coverage and average set size. If we desire coverage guarantees, then we recommend the repeated subsampling approach.
Cites work
- scientific article; zbMATH DE number 2168212 (Why is no real title available?)
- A tutorial on conformal prediction
- Combining p-values via averaging
- Distribution-Free Prediction Sets
- Distribution-free Prediction Bands for Non-parametric Regression
- Distribution-free predictive inference for regression
- Large‐scale simultaneous inference under dependence
- Least Ambiguous Set-Valued Classifiers With Bounded Error Levels
- Meta-analysis with fixed, unknown, study-specific parameters
- Posterior predictive \(p\)-values
- Predictive inference using latent variables with covariates
- Predictive inference with the jackknife+
- Random variables with maximum sums
- Random-Effects Models for Longitudinal Data
- Standard Errors of Prediction in Generalized Linear Mixed Models
- The limits of distribution-free conditional predictive inference
Cited in
(3)
This page was built for publication: Distribution-Free Prediction Sets for Two-Layer Hierarchical Models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6144764)
