Saved in:
Bibliographic Details
Main Authors: van Erp, Bart, de Vries, Bert
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.10395
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917803670372352
author van Erp, Bart
de Vries, Bert
author_facet van Erp, Bart
de Vries, Bert
contents This paper proposes improvements over earlier work by Nazareth and Blei (2022) for estimating the depth of Bayesian neural networks. Here, we propose a discrete truncated normal distribution over the network depth to independently learn its mean and variance. Posterior distributions are inferred by minimizing the variational free energy, which balances the model complexity and accuracy. Our method improves test accuracy on the spiral data set and reduces the variance in posterior depth estimates.
format Preprint
id arxiv_https___arxiv_org_abs_2410_10395
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Improved Depth Estimation of Bayesian Neural Networks
van Erp, Bart
de Vries, Bert
Machine Learning
This paper proposes improvements over earlier work by Nazareth and Blei (2022) for estimating the depth of Bayesian neural networks. Here, we propose a discrete truncated normal distribution over the network depth to independently learn its mean and variance. Posterior distributions are inferred by minimizing the variational free energy, which balances the model complexity and accuracy. Our method improves test accuracy on the spiral data set and reduces the variance in posterior depth estimates.
title Improved Depth Estimation of Bayesian Neural Networks
topic Machine Learning
url https://arxiv.org/abs/2410.10395