Computer Science > Machine Learning
[Submitted on 16 Jun 2009 (v1), revised 12 Oct 2009 (this version, v3), latest version 8 Nov 2010 (v6)]
Title:Entropy Message Passing Algorithm
View PDFAbstract: Message passing over factor graph can be considered as generalization of many well known algorithms for efficient marginalization of multivariate function. A specific instance of the algorithm is obtained by choosing an appropriate commutative semiring for the range of the function to be marginalized. Some examples are Viterbi algorithm, obtained on max-product semiring and forward-backward algorithm obtained on sum-product semiring. In this paper, Entropy Message Passing algorithm (EMP) is developed. It operates over entropy semiring, previously introduced in automata theory. It is shown how EMP extends the use of message passing over factor graphs to probabilistic model algorithms such as Expectation Maximization algorithm, gradient methods and computation of model entropy, unifying the work of different authors.
Submission history
From: Velimir Ilic [view email][v1] Tue, 16 Jun 2009 10:58:50 UTC (55 KB)
[v2] Tue, 16 Jun 2009 23:03:13 UTC (31 KB)
[v3] Mon, 12 Oct 2009 12:48:20 UTC (31 KB)
[v4] Thu, 13 May 2010 16:40:01 UTC (251 KB)
[v5] Sat, 15 May 2010 22:23:39 UTC (64 KB)
[v6] Mon, 8 Nov 2010 01:37:37 UTC (61 KB)
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.