Convexifying the Bethe Free Energy
arXiv:1205.2624
Abstract
The introduction of loopy belief propagation (LBP) revitalized the application of graphical models in many domains. Many recent works present improvements on the basic LBP algorithm in an attempt to overcome convergence and local optima problems. Notable among these are convexified free energy approximations that lead to inference procedures with provable convergence and quality properties. However, empirically LBP still outperforms most of its convex variants in a variety of settings, as we also demonstrate here. Motivated by this fact we seek convexified free energies that directly approximate the Bethe free energy. We show that the proposed approximations compare favorably with state-of-the art convex free energy approximations.
Appears in Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence (UAI2009)
References in corpus (5)
- Loopy Belief Propagation for Approximate Inference: An Empirical Study
- Belief Optimization for Binary Networks: A Stable Alternative to Loopy Belief Propagation
- Approximate Inference and Constrained Optimization
- Convergent Message-Passing Algorithms for Inference over General Graphs with Convex Free Energies
- Convergent Propagation Algorithms via Oriented Trees