The EM algorithm and the Laplace Approximation
Machine Learning
2014-01-27 v1
Abstract
The Laplace approximation calls for the computation of second derivatives at the likelihood maximum. When the maximum is found by the EM-algorithm, there is a convenient way to compute these derivatives. The likelihood gradient can be obtained from the EM-auxiliary, while the Hessian can be obtained from this gradient with the Pearlmutter trick.
Cite
@article{arxiv.1401.6276,
title = {The EM algorithm and the Laplace Approximation},
author = {Niko Brümmer},
journal= {arXiv preprint arXiv:1401.6276},
year = {2014}
}