Repository logo
  • English
  • Italiano
  • Log In
    Have you forgotten your password?
Repository logo
Repository logo
  • Archive
  • Series/Journals
  • EUT
  • Events
  • Statistics
  • English
  • Italiano
  • Log In
    Have you forgotten your password?
  1. Home
  2. EUT Edizioni Università di Trieste
  3. Collane
  4. DEAMS Research Paper Series
  5. DEAMS Research Paper Series 2020, 2
  6. Streamlined Variational Inference for Linear Mixed Models with Crossed Random Effects
 
  • Details
  • Metrics
Options

Streamlined Variational Inference for Linear Mixed Models with Crossed Random Effects

Menictas, Marianne
•
Di Credico, Gioia
•
Wand, Matt P.
2020
Loading...
Thumbnail Image
http://hdl.handle.net/10077/30908
  • Book

e-ISBN
978-88-5511-190-4
Abstract
We derive streamlined mean field variational Bayes algorithms for fitting linear mixed models with crossed random effects. In the most general situation, where the dimensions of the crossed groups are arbitrarily large, streamlining is hindered by lack of sparseness in the underlying least squares system. Because of this fact we also consider a hierarchy of relaxations of the mean field product restriction. The least stringent product restriction delivers a high degree of inferential accuracy. However, this accuracy must be mitigated against its higher storage and computing demands. Faster sparse storage and computing alternatives are also provided, but come with the price of diminished inferential accu-racy. This article provides full algorithmic details of three variational inference strategies, presents detailed empirical results on their pros and cons and, thus, guides the users on their choice of variational inference approach depending on the problem size and computing resources.
Series
DEAMS Research Paper Series, 2020, 2
Subjects
  • Mean field variational...

  • Item response theory

  • Rasch analysis

  • Scalable statistical ...

  • Sparse least squares ...

Publisher
EUT Edizioni Università di Trieste
Source
Marianne Menictas, Gioia Di Credico, Matt P. Wand, "Streamlined Variational Inference for Linear Mixed Models with Crossed Random Effects", Trieste, EUT Edizioni Università di Trieste, 2020
Languages
en
Rights
Attribution-NonCommercial-NoDerivatives 4.0 Internazionale
Licence
http://creativecommons.org/licenses/by-nc-nd/4.0/
File(s)
Loading...
Thumbnail Image
Download
Name

DEAMS_RP_2020_2_Menictas_DiCredico_Wand.pdf

Format

Adobe PDF

Size

1.08 MB

Loading...
Thumbnail Image
Download
Name

DEAMS_RP_2020_2_Front_cover.jpg

Format

JPEG

Size

213.03 KB

Indexed by

 Info

Open Access Policy

Share/Save

 Contacts

EUT Edizioni Università di Trieste

OpenstarTs

 Link

Wiki OpenAcces

Archivio Ricerca ArTS

Built with DSpace-CRIS software - Extension maintained and optimized by 4Science

  • Cookie settings
  • Privacy policy
  • End User Agreement
  • Send Feedback