CUED Publications database

Clamping Improves TRW and Mean Field Approximations

Weller, A and Domke, J Clamping Improves TRW and Mean Field Approximations. (Unpublished)

Full text not available from this repository.


We examine the effect of clamping variables for approximate inference in undirected graphical models with pairwise relationships and discrete variables. For any number of variable labels, we demonstrate that clamping and summing approximate sub-partition functions can lead only to a decrease in the partition function estimate for TRW, and an increase for the naive mean field method, in each case guaranteeing an improvement in the approximation and bound. We next focus on binary variables, add the Bethe approximation to consideration and examine ways to choose good variables to clamp, introducing new methods. We show the importance of identifying highly frustrated cycles, and of checking the singleton entropy of a variable. We explore the value of our methods by empirical analysis and draw lessons to guide practitioners.

Item Type: Article
Uncontrolled Keywords: cs.LG cs.LG cs.AI stat.ML
Divisions: Div F > Computational and Biological Learning
Depositing User: Cron Job
Date Deposited: 15 Nov 2017 20:14
Last Modified: 18 Feb 2021 18:14