Approximate inference using DC programming for collective graphical models

Collective graphical models (CGMs) provide a framework for reasoning about a population of independent and identically distributed individuals when only noisy and aggregate observations are given. Previous approaches for inference in CGMs work on a junction-tree representation, thereby highly limiti...

Full description

Saved in:
Bibliographic Details
Main Authors: NGUYEN, Duc Thien, Akshat KUMAR, LAU, Hoong Chuin, SHELDON, Daniel
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2016
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/3400
https://ink.library.smu.edu.sg/context/sis_research/article/4401/viewcontent/ApproximateInterference.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-4401
record_format dspace
spelling sg-smu-ink.sis_research-44012020-04-08T00:58:47Z Approximate inference using DC programming for collective graphical models NGUYEN, Duc Thien Akshat KUMAR, LAU, Hoong Chuin SHELDON, Daniel Collective graphical models (CGMs) provide a framework for reasoning about a population of independent and identically distributed individuals when only noisy and aggregate observations are given. Previous approaches for inference in CGMs work on a junction-tree representation, thereby highly limiting their scalability. To remedy this, we show how the Bethe entropy approximation naturally arises for the inference problem in CGMs. We reformulate the resulting optimization problem as a difference-of-convex functions program that can capture different types of CGM noise models. Using the concave-convex procedure, we then develop a scalable message-passing algorithm. Empirically, our approach is highly scalable and accurate for large graphs, more than an order-of-magnitude faster than a generic optimization solver, and is guaranteed to converge unlike the previous message-passing approach NLBP that fails in several loopy graphs. 2016-05-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/3400 https://ink.library.smu.edu.sg/context/sis_research/article/4401/viewcontent/ApproximateInterference.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University Approximate inference Concave-convex procedure Difference of convex functions Entropy approximations Generic optimization Inference problem Message passing algorithm Optimization problems Artificial Intelligence and Robotics Computer Sciences Operations Research, Systems Engineering and Industrial Engineering
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic Approximate inference
Concave-convex procedure
Difference of convex functions
Entropy approximations
Generic optimization
Inference problem
Message passing algorithm
Optimization problems
Artificial Intelligence and Robotics
Computer Sciences
Operations Research, Systems Engineering and Industrial Engineering
spellingShingle Approximate inference
Concave-convex procedure
Difference of convex functions
Entropy approximations
Generic optimization
Inference problem
Message passing algorithm
Optimization problems
Artificial Intelligence and Robotics
Computer Sciences
Operations Research, Systems Engineering and Industrial Engineering
NGUYEN, Duc Thien
Akshat KUMAR,
LAU, Hoong Chuin
SHELDON, Daniel
Approximate inference using DC programming for collective graphical models
description Collective graphical models (CGMs) provide a framework for reasoning about a population of independent and identically distributed individuals when only noisy and aggregate observations are given. Previous approaches for inference in CGMs work on a junction-tree representation, thereby highly limiting their scalability. To remedy this, we show how the Bethe entropy approximation naturally arises for the inference problem in CGMs. We reformulate the resulting optimization problem as a difference-of-convex functions program that can capture different types of CGM noise models. Using the concave-convex procedure, we then develop a scalable message-passing algorithm. Empirically, our approach is highly scalable and accurate for large graphs, more than an order-of-magnitude faster than a generic optimization solver, and is guaranteed to converge unlike the previous message-passing approach NLBP that fails in several loopy graphs.
format text
author NGUYEN, Duc Thien
Akshat KUMAR,
LAU, Hoong Chuin
SHELDON, Daniel
author_facet NGUYEN, Duc Thien
Akshat KUMAR,
LAU, Hoong Chuin
SHELDON, Daniel
author_sort NGUYEN, Duc Thien
title Approximate inference using DC programming for collective graphical models
title_short Approximate inference using DC programming for collective graphical models
title_full Approximate inference using DC programming for collective graphical models
title_fullStr Approximate inference using DC programming for collective graphical models
title_full_unstemmed Approximate inference using DC programming for collective graphical models
title_sort approximate inference using dc programming for collective graphical models
publisher Institutional Knowledge at Singapore Management University
publishDate 2016
url https://ink.library.smu.edu.sg/sis_research/3400
https://ink.library.smu.edu.sg/context/sis_research/article/4401/viewcontent/ApproximateInterference.pdf
_version_ 1770573159920041984