Bayesian Inference with Measurement Error
Bayesian inference with measurement error extends the standard Bayesian framework to situations where one or more covariates or outcomes are observed with noise or misclassification. By treating the true unobserved values as latent variables and assigning them priors, the model jointly estimates the true exposure distribution and the structural parameters of interest, propagating all uncertainty through the posterior.
Kilderegistrering
Citater kopieret ordret fra metodens kilderegistrering. Ingen påstandsniveauverifikation er udledt heraf.
- Carroll, R. J., Ruppert, D., Stefanski, L. A., & Crainiceanu, C. M. (2006). Measurement Error in Nonlinear Models: A Modern Perspective (2nd ed.). Chapman & Hall/CRC. · ISBN 978-1584886433
- Richardson, S., & Gilks, W. R. (1993). A Bayesian approach to measurement error problems in epidemiology using conditional independence models. American Journal of Epidemiology, 138(6), 430–442. · DOI 10.1093/oxfordjournals.aje.a116875
Kuraterede påstande
Påstande gemt i bevis-loggen, hver med sin egen vurdering.
Denne visning opfinder ikke en påstandsvurdering, når loggen ingen har.
Relaterede metoder
Genereret fra metodegrafen og vist som maskinelt foreslåede relationer — ingen bevispåstand er udledt.