Accurate logistic variational message passing: algebraic and numerical details

Nolan, TH; Wand, MP

Accurate logistic variational message passing: algebraic and numerical details

Nolan, TH Wand, MP

Permalink

Publication Type:: Journal Article
Citation:: Stat, 2017, 6 (1), pp. 102 - 112
Issue Date:: 2017-01-01

Closed Access

	Filename	Description	Size
	Nolan_et_al-2017-Stat.pdf	Submitted Version	801.76 kB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Nolan, TH	en_US
dc.contributor.author	Wand, MP https://orcid.org/0000-0003-2555-896X	en_US
dc.date.issued	2017-01-01	en_US
dc.identifier.citation	Stat, 2017, 6 (1), pp. 102 - 112	en_US
dc.identifier.uri	http://hdl.handle.net/10453/111915
dc.description.abstract	Copyright © 2017 John Wiley & Sons, Ltd. We provide full algebraic and numerical details required for fitting accurate logistic likelihood regression-type models via variational message passing with factor graph fragments. Existing methodology of this type involves the Jaakkola–Jordan device, which is prone to poor accuracy. We examine two alternatives: the Saul–Jordan tilted bound device and conjugacy enforcement via multivariate normal prespecification of a key message. Both of these approaches appear in related literature. Our contributions facilitate immediate implementation within variational message passing schemes. Copyright © 2017 John Wiley & Sons, Ltd.	en_US
dc.relation	http://purl.org/au-research/grants/arc/DP140100441
dc.relation.ispartof	Stat	en_US
dc.relation.isbasedon	10.1002/sta4.139	en_US
dc.title	Accurate logistic variational message passing: algebraic and numerical details	en_US
dc.type	Journal Article
utslib.citation.volume	1	en_US
utslib.citation.volume	6	en_US
utslib.for	0104 Statistics	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Science
pubs.organisational-group	/University of Technology Sydney/Faculty of Science/School of Mathematical and Physical Sciences
utslib.copyright.status	closed_access
pubs.issue	1	en_US
pubs.publication-status	Published	en_US
pubs.volume	6	en_US

Abstract:

Copyright © 2017 John Wiley & Sons, Ltd. We provide full algebraic and numerical details required for fitting accurate logistic likelihood regression-type models via variational message passing with factor graph fragments. Existing methodology of this type involves the Jaakkola–Jordan device, which is prone to poor accuracy. We examine two alternatives: the Saul–Jordan tilted bound device and conjugacy enforcement via multivariate normal prespecification of a key message. Both of these approaches appear in related literature. Our contributions facilitate immediate implementation within variational message passing schemes. Copyright © 2017 John Wiley & Sons, Ltd.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/111915