Tractable Variational Structures for Approximating Graphical Models
By David Barber, Wim Wiegerinck. 1998
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- David M. Blei
- Sergey Levine
- Yoshua Bengio
- Michael I. Jordan
- Ruslan Salakhutdinov
- Trevor Darrell
- Geoffrey E. Hinton
- Zhe Gan
Citations in the corpus
- Variational Inference: A Review for Statisticians (2016, background): “As wementioned previously, structured variational inference has its roots in the early days of the method (Saul and Jordan 1996; Barber and Wiegerinck 1999).”
- Variational Cumulant Expansions for Intractable Distributions (2011, methodology): “These tractable distributions have recently been exploited in the variational method to approximate intractable distributions, see for example (Saul, Jaakkola, & Jordan, 1996; Jaakkola, 1997; Barber & Wiegerinck, 1999; Jordan et al., 1998).”
- A family of algorithms for approximate Bayesian inference (2001, methodology): “…approximations have previously been used in ADF by Frey et al. (2000) and in the context of variational bounds by Ghahramani & Jordan (1997), Barber & Wiegerinck (1998), and references therein. x 2 x 1 x 4 x 3 x 5 x 6 Figure 4-5: Belief propagation can also be improved by using a Markov…”
- Graphical Models, Exponential Families, and Variational Inference (2008, methodology): “This structured mean field approach was first proposed by Saul and Jordan [204], and further developed by various researchers [9, 254, 117].”
- An Auxiliary Variational Method (2004, methodology): “Variational approximations have been extensively used in physics and engineering and more recently applied to approximate inference and learning in intractable graphical models (see e.g. Jordan et al. (1998); Barber and Wiegerinck (1998); Wainwright et al. (2002)).”
- Bayesian reasoning and machine learning (2012, unknown): “5 Structured variational approximation One can extend the factorised KL variational approximation by using non-factorised q(x)[239, 24].”
- A Tighter Bound for Graphical Models (2001, methodology): “To assess the error made by the various approaches, we use the same toy problem as in [2] and [10].”