The effect of ambiguous prior knowledge on Bayesian model parameter inference and prediction
Lists
Only you see this
Abstract
Highlights•There is often ambiguity about the choice of Bayesian prior probability distribution.•The Density Ratio Class represents such ambiguity using sets of densities.•We show that this class is invariant under inference, marginalization, and prediction.•Such properties are conceptually satisfying and enable computational efficiency.•We demonstrate concepts and new algorithms using a simple ecological model.
BibTeX
@inproceedings{rinderknecht2014effect,
author = {Simon L. Rinderknecht and Carlo Albert and Mark E. Borsuk and Nele Schuwirth and Hans R. Künsch and Peter Reichert},
title = {{The effect of ambiguous prior knowledge on Bayesian model parameter inference and prediction}},
booktitle = {Environ. Model. Softw},
year = {2014},
url = {https://openreview.net/forum?id=XLVTWLc22s},
}Loading related content...