DeepAI AI Chat
Log In Sign Up

Dual Decomposition from the Perspective of Relax, Compensate and then Recover

by   Arthur Choi, et al.

Relax, Compensate and then Recover (RCR) is a paradigm for approximate inference in probabilistic graphical models that has previously provided theoretical and practical insights on iterative belief propagation and some of its generalizations. In this paper, we characterize the technique of dual decomposition in the terms of RCR, viewing it as a specific way to compensate for relaxed equivalence constraints. Among other insights gathered from this perspective, we propose novel heuristics for recovering relaxed equivalence constraints with the goal of incrementally tightening dual decomposition approximations, all the way to reaching exact solutions. We also show empirically that recovering equivalence constraints can sometimes tighten the corresponding approximation (and obtaining exact results), without increasing much the complexity of inference.


page 1

page 2

page 3

page 4


Lifted Relax, Compensate and then Recover: From Approximate to Exact Lifted Probabilistic Inference

We propose an approach to lifted approximate inference for first-order p...

Belief Propagation for Structured Decision Making

Variational inference algorithms such as belief propagation have had tre...

Maximum Persistency via Iterative Relaxed Inference with Graphical Models

We consider the NP-hard problem of MAP-inference for undirected discrete...

First-Order Decomposition Trees

Lifting attempts to speed up probabilistic inference by exploiting symme...

Efficiently Searching for Frustrated Cycles in MAP Inference

Dual decomposition provides a tractable framework for designing algorith...

Relaxed Scheduling for Scalable Belief Propagation

The ability to leverage large-scale hardware parallelism has been one of...