Choice of Basis for Laplace Approximation
Abstract
Maximum a posteriori optimization of parameters and the Laplace approximation for the marginal likelihood are both basis-dependent methods. This note compares two choices of basis for models parameterized by probabilities, showing that it is possible to improve on the traditional choice, the probability simplex, by transforming to the 'softmax' basis.
Cite
Text
MacKay. "Choice of Basis for Laplace Approximation." Machine Learning, 1998. doi:10.1023/A:1007558615313Markdown
[MacKay. "Choice of Basis for Laplace Approximation." Machine Learning, 1998.](https://mlanthology.org/mlj/1998/mackay1998mlj-choice/) doi:10.1023/A:1007558615313BibTeX
@article{mackay1998mlj-choice,
title = {{Choice of Basis for Laplace Approximation}},
author = {MacKay, David J. C.},
journal = {Machine Learning},
year = {1998},
pages = {77-86},
doi = {10.1023/A:1007558615313},
volume = {33},
url = {https://mlanthology.org/mlj/1998/mackay1998mlj-choice/}
}