Predicting future responses based on possibly mis-specified working models

DOI10.1093/BIOMET/ASM078MaRDI QIDQ3631481zbMATH OpenOpenAlexFDO

Authors Tianxi Cai, Lu Tian, Scott D. Solomon, L. J. Wei

Publication date 10 June 2009

Published in Biometrika (Search for Journal in Brave)

Full work available at URL https://biostats.bepress.com/harvardbiostat/paper48

zbMATH Keywords

\(K\)-fold crossvalidationheterogeneous regressionmis-specified regression modeloptimal prediction regionprediction error rate

Mathematics Subject Classification ID

Applications of statistics to biology and medical sciences; meta analysis (62P10)

Recommendations

• Model imperfection and predicting predictability
• Response prediction in mixed effects models
• Models for understanding versus models for prediction
• Predictive inference for future responses from two component systems
• Model misspecification tests, model building and predictability in complex systems
• Prediction in several conventional contexts
• Modeling and predicting extrapolated probabilities with outlooks (With discussion)
• Reflections on `Adaptive prediction'
• Predicting the future. Completing models of observed complex systems

Cited in

(7)

• Applications of hyperellipsoidal prediction regions
• Comparing six shrinkage estimators with large sample theory and asymptotically optimal prediction intervals
• GMM and misspecification correction for misspecified models with diverging number of parameters
• A new prediction interval for binomial random variable based on inferential models
• Sparse kernel machine regression for ordinal outcomes
• Prediction intervals for GLMs, GAMs, and some survival regression models
• A nonparametric distribution-free test for serial independence of errors

This page was built for publication: Predicting future responses based on possibly mis-specified working models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3631481)