Streamlining Prediction in Bayesian Deep Learning

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Li, Rui, Klasson, Marcus, Solin, Arno, Trapp, Martin
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912495477719040
author Li, Rui
Klasson, Marcus
Solin, Arno
Trapp, Martin
author_facet Li, Rui
Klasson, Marcus
Solin, Arno
Trapp, Martin
contents The rising interest in Bayesian deep learning (BDL) has led to a plethora of methods for estimating the posterior distribution. However, efficient computation of inferences, such as predictions, has been largely overlooked with Monte Carlo integration remaining the standard. In this work we examine streamlining prediction in BDL through a single forward pass without sampling. For this we use local linearisation on activation functions and local Gaussian approximations at linear layers. Thus allowing us to analytically compute an approximation to the posterior predictive distribution. We showcase our approach for both MLP and transformers, such as ViT and GPT-2, and assess its performance on regression and classification tasks. Open-source library: https://github.com/AaltoML/SUQ
format Preprint
id arxiv_https___arxiv_org_abs_2411_18425
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Streamlining Prediction in Bayesian Deep Learning
Li, Rui
Klasson, Marcus
Solin, Arno
Trapp, Martin
Machine Learning
The rising interest in Bayesian deep learning (BDL) has led to a plethora of methods for estimating the posterior distribution. However, efficient computation of inferences, such as predictions, has been largely overlooked with Monte Carlo integration remaining the standard. In this work we examine streamlining prediction in BDL through a single forward pass without sampling. For this we use local linearisation on activation functions and local Gaussian approximations at linear layers. Thus allowing us to analytically compute an approximation to the posterior predictive distribution. We showcase our approach for both MLP and transformers, such as ViT and GPT-2, and assess its performance on regression and classification tasks. Open-source library: https://github.com/AaltoML/SUQ
title Streamlining Prediction in Bayesian Deep Learning
topic Machine Learning
url https://arxiv.org/abs/2411.18425