Skip to content

Commit

Permalink
Merge pull request #188 from biaslab/dev-pruningBNN
Browse files Browse the repository at this point in the history
Update BNN pruning paper
  • Loading branch information
Nimrais authored Jan 2, 2024
2 parents 3f3f8bc + e72454a commit f8d0d9e
Showing 1 changed file with 5 additions and 5 deletions.
10 changes: 5 additions & 5 deletions content/publication/principled_pruning_BNN.md
Original file line number Diff line number Diff line change
@@ -1,19 +1,19 @@
+++
abstract = "Bayesian model reduction provides an efficient approach for comparing the performance of all nested sub-models of a model, without re-evaluating any of these sub-models. Until now, Bayesian model reduction has been applied mainly in the computational neuroscience community. In this paper, we formulate and apply Bayesian model reduction to perform principled pruning of Bayesian neural networks, based on variational free energy minimization. This novel parameter pruning scheme solves the shortcomings of many current state-of-the-art pruning methods that are used by the signal processing community. The proposed approach has a clear stopping criterion and minimizes the same objective that is used during training. Next to these theoretical benefits, our experiments indicate better model performance in comparison to state-of-the-art pruning schemes."
date = "2022-10-17"
abstract = "Bayesian model reduction provides an efficient approach for comparing the performance of all nested sub-models of a model, without re-evaluating any of these sub-models. Until now, Bayesian model reduction has been applied mainly in the computational neuroscience community on simple models. In this paper, we formulate and apply Bayesian model reduction to perform principled pruning of Bayesian neural networks, based on variational free energy minimization. Direct application of Bayesian model reduction, however, gives rise to approximation errors. Therefore, a novel iterative pruning algorithm is presented to alleviate the problems arising with naive Bayesian model reduction, as supported experimentally on the publicly available UCI datasets for different inference algorithms. This novel parameter pruning scheme solves the shortcomings of current state-of-the-art pruning methods that are used by the signal processing community. The proposed approach has a clear stopping criterion and minimizes the same objective that is used during training. Next to these benefits, our experiments indicate better model performance in comparison to state-of-the-art pruning schemes."
date = "2023-11-29"
image = ""
image_preview = ""
math = false
publication = "arXiv preprint"
publication = "IEEE Open Journal of Signal Processing"
to_be_published = false
selected = false
title = "Principled Pruning of Bayesian Neural Networks through Variational Free Energy Minimization"
url_code = "https://github.com/biaslab/principledpruningBNN"
url_dataset = ""
url_pdf = "https://arxiv.org/pdf/2210.09134.pdf"
url_pdf = "https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10334001"
url_project = ""
url_video = ""
url_custom = [{name="arXiv", url = "https://arxiv.org/abs/2210.09134"}]
url_custom = [{name = "IEEE", url = "https://ieeexplore.ieee.org/document/10334001"}, {name = "arXiv", url = "https://arxiv.org/abs/2210.09134"}]

[[authors]]
id = "jim"
Expand Down

0 comments on commit f8d0d9e

Please sign in to comment.