Bayesian learning via neural Schrödinger–Föllmer flows

Francisco Vargas, Andrius Ovsianas, David Fernandes, Mark Girolami, Neil D. Lawrence, Nikolas Nüsken

Research output: Contribution to journalArticlepeer-review

1 Citation (SciVal)

Abstract

In this work we explore a new framework for approximate Bayesian inference in large datasets based on stochastic control. We advocate stochastic control as a finite time and low variance alternative to popular steady-state methods such as stochastic gradient Langevin dynamics. Furthermore, we discuss and adapt the existing theoretical guarantees of this framework and establish connections to already existing VI routines in SDE-based models.

Original languageEnglish
Article number3
JournalStatistics and Computing
Volume33
Issue number1
Early online date23 Nov 2022
DOIs
Publication statusPublished - 28 Feb 2023

Bibliographical note

Funding Information:
Francisco Vargas is Funded by Huawei Technologies Co. This research has been partially funded by Deutsche Forschungsgemeinschaft (DFG) through the grant CRC 1114 ‘Scaling Cascades in Complex Systems’ (project A02, project number 235221301). Andrius Ovsianas is funded by EPSRC iCASE Award EP/T517677/1. Mark Girolami is supported by a Royal Academy of Engineering Research Chair, and EPSRC grants EP/T000414/1, EP/R018413/2, EP/P020720/2, EP/R034710/1, EP/R004889/1.

Funding Information:
Francisco Vargas is Funded by Huawei Technologies Co. This research has been partially funded by Deutsche Forschungsgemeinschaft (DFG) through the grant CRC 1114 ‘Scaling Cascades in Complex Systems’ (project A02, project number 235221301). Andrius Ovsianas is funded by EPSRC iCASE Award EP/T517677/1. Mark Girolami is supported by a Royal Academy of Engineering Research Chair, and EPSRC grants EP/T000414/1, EP/R018413/2, EP/P020720/2, EP/R034710/1, EP/R004889/1.

Keywords

  • Bayesian deep learning
  • Bayesian inference
  • Föllmer drift
  • Schrödinger bridge problem
  • Stochastic control

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Computational Theory and Mathematics

Fingerprint

Dive into the research topics of 'Bayesian learning via neural Schrödinger–Föllmer flows'. Together they form a unique fingerprint.

Cite this