Gauging variational inference
From MaRDI portal
Abstract: Computing partition function is the most important statistical inference task arising in applications of Graphical Models (GM). Since it is computationally intractable, approximate methods have been used to resolve the issue in practice, where mean-field (MF) and belief propagation (BP) are arguably the most popular and successful approaches of a variational type. In this paper, we propose two new variational schemes, coined Gauged-MF (G-MF) and Gauged-BP (G-BP), improving MF and BP, respectively. Both provide lower bounds for the partition function by utilizing the so-called gauge transformation which modifies factors of GM while keeping the partition function invariant. Moreover, we prove that both G-MF and G-BP are exact for GMs with a single loop of a special structure, even though the bare MF and BP perform badly in this case. Our extensive experiments, on complete GMs of relatively small size and on large GM (up-to 300 variables) confirm that the newly proposed algorithms outperform and generalize MF and BP.
Recommendations
Cites work
- A New Class of Upper Bounds on the Log Partition Function
- Belief propagation and loop series on planar graphs
- Codes on graphs: normal realizations
- Constructing Free-Energy Approximations and Generalized Belief Propagation Algorithms
- Graphical models, exponential families, and variational inference
- Holographic Algorithms
- scientific article; zbMATH DE number 3174791 (Why is no real title available?)
- scientific article; zbMATH DE number 1273988 (Why is no real title available?)
- scientific article; zbMATH DE number 6317138 (Why is no real title available?)
- Information, Physics, and Computation
- Learning low-level vision
- Loop calculus in statistical physics and information science
- Loop series for discrete statistical models on graphs
- Normal Factor Graphs and Holographic Transformations
- On Ising's model of ferromagnetism
- On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming
- Polynomial-Time Approximation Algorithms for the Ising Model
- Statistical field theory. With a foreword by David Pines
- Statistical theory of superlattices
- Tree-based reparameterization framework for analysis of sum-product and related algorithms
Cited in
(14)- Variational inference at glacier scale
- The computational asymptotics of Gaussian variational inference and the Laplace approximation
- Conditionally structured variational Gaussian approximation with importance weights
- Variational inference for latent variables and uncertain inputs in Gaussian processes
- scientific article; zbMATH DE number 6377992 (Why is no real title available?)
- On the properties of variational approximations of Gibbs posteriors
- The Variational Gaussian Approximation Revisited
- Realizing Active Inference in Variational Message Passing: The Outcome-Blind Certainty Seeker
- Optimizing Variational Representations of Divergences and Accelerating Their Statistical Estimation
- scientific article; zbMATH DE number 7307467 (Why is no real title available?)
- Improving variational methods via pairwise linear response identities
- Bucket renormalization for approximate inference
- Gauges, loops, and polynomials for partition functions of graphical models
- Amortized Variational Inference: A Systematic Review
This page was built for publication: Gauging variational inference
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5854117)
