Beyond Dirichlet-based Models: When Bayesian Neural Networks Meet Evidential Deep Learning

Hanjing Wang, Qiang Ji
Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence, PMLR 244:3643-3665, 2024.

Abstract

Bayesian neural networks (BNNs) excel in uncertainty quantification (UQ) by estimating the posterior distribution of model parameters, yet face challenges due to the high computational demands of Bayesian inference. Evidential deep learning methods address this by treating target distribution parameters as random variables with a learnable conjugate distribution, enabling efficient UQ. However, there’s debate over whether these methods can accurately estimate epistemic uncertainty due to their single-network, sampling-free nature. In this paper, we combine the strengths of both approaches by distilling BNN knowledge into a Dirichlet-based model, endowing it with a Bayesian perspective and theoretical guarantees. Additionally, we introduce two enhancements to further improve the integration of Bayesian UQ with Dirichlet-based models. To relax the heavy computational load with BNNs, we introduce a self-regularized training strategy using Laplacian approximation (LA) for self-distillation. To alleviate the conjugate prior assumption, we employ an expressive normalizing flow for refining the model in a post-processing manner, where a few training iterations can enhance model performance. The experimental results have demonstrated the effectiveness of our proposed methods in both UQ accuracy and robustness.

Cite this Paper

BibTeX
@InProceedings{pmlr-v244-wang24g, title = {Beyond Dirichlet-based Models: When Bayesian Neural Networks Meet Evidential Deep Learning}, author = {Wang, Hanjing and Ji, Qiang}, booktitle = {Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence}, pages = {3643--3665}, year = {2024}, editor = {Kiyavash, Negar and Mooij, Joris M.}, volume = {244}, series = {Proceedings of Machine Learning Research}, month = {15--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v244/main/assets/wang24g/wang24g.pdf}, url = {https://proceedings.mlr.press/v244/wang24g.html}, abstract = {Bayesian neural networks (BNNs) excel in uncertainty quantification (UQ) by estimating the posterior distribution of model parameters, yet face challenges due to the high computational demands of Bayesian inference. Evidential deep learning methods address this by treating target distribution parameters as random variables with a learnable conjugate distribution, enabling efficient UQ. However, there’s debate over whether these methods can accurately estimate epistemic uncertainty due to their single-network, sampling-free nature. In this paper, we combine the strengths of both approaches by distilling BNN knowledge into a Dirichlet-based model, endowing it with a Bayesian perspective and theoretical guarantees. Additionally, we introduce two enhancements to further improve the integration of Bayesian UQ with Dirichlet-based models. To relax the heavy computational load with BNNs, we introduce a self-regularized training strategy using Laplacian approximation (LA) for self-distillation. To alleviate the conjugate prior assumption, we employ an expressive normalizing flow for refining the model in a post-processing manner, where a few training iterations can enhance model performance. The experimental results have demonstrated the effectiveness of our proposed methods in both UQ accuracy and robustness.} }
Endnote
%0 Conference Paper %T Beyond Dirichlet-based Models: When Bayesian Neural Networks Meet Evidential Deep Learning %A Hanjing Wang %A Qiang Ji %B Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2024 %E Negar Kiyavash %E Joris M. Mooij %F pmlr-v244-wang24g %I PMLR %P 3643--3665 %U https://proceedings.mlr.press/v244/wang24g.html %V 244 %X Bayesian neural networks (BNNs) excel in uncertainty quantification (UQ) by estimating the posterior distribution of model parameters, yet face challenges due to the high computational demands of Bayesian inference. Evidential deep learning methods address this by treating target distribution parameters as random variables with a learnable conjugate distribution, enabling efficient UQ. However, there’s debate over whether these methods can accurately estimate epistemic uncertainty due to their single-network, sampling-free nature. In this paper, we combine the strengths of both approaches by distilling BNN knowledge into a Dirichlet-based model, endowing it with a Bayesian perspective and theoretical guarantees. Additionally, we introduce two enhancements to further improve the integration of Bayesian UQ with Dirichlet-based models. To relax the heavy computational load with BNNs, we introduce a self-regularized training strategy using Laplacian approximation (LA) for self-distillation. To alleviate the conjugate prior assumption, we employ an expressive normalizing flow for refining the model in a post-processing manner, where a few training iterations can enhance model performance. The experimental results have demonstrated the effectiveness of our proposed methods in both UQ accuracy and robustness.
APA
Wang, H. & Ji, Q.. (2024). Beyond Dirichlet-based Models: When Bayesian Neural Networks Meet Evidential Deep Learning. Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 244:3643-3665 Available from https://proceedings.mlr.press/v244/wang24g.html.

Related Material