Fetching the paper…

Reading the bibliography…

2014

Mean-Field Networks

Li, Yujia, Zemel, Richard

Original
PDF
Post on X

Understand

The mean field algorithm is a widely used approximate inference algorithm for graphical models whose exact inference is intractable.

In each iteration of mean field, the approximate marginals for each variable are updated by getting information from the neighbors.
This process can be equivalently converted into a feedforward network, with each layer representing one iteration of mean field and with tied weights on all layers.
This conversion enables a few natural extensions, e.g.

Mean-Field Networks · Around

Built on

The helmholtz machine
Dayan, P., Hinton, G.E., Neal, R.M., and Zemel, R.S · 1995
Earlier work this paper cites.
Estimating the wrong graphical model: Benefits in the computation-limited setting
Original
Wainwright, Martin J · 2006
Earlier work this paper cites.
Supervised learning of image restoration with convolutional networks
Jain, Viren et al · 2007
Earlier work this paper cites.

Similar

Structured learning with approximate inference
Kulesza, Alex and Pereira, Fernando · 2007
Cited alongside, same era.
Parameter learning with truncated message-passing
Domke, Justin · 2011
Cited alongside, same era.
Empirical risk minimization of graphical model parameters given approximate inference, decoding, and model structure
Stoyanov, Veselin, Ropson, Alexander, and Eisner, Jason · 2011
Cited alongside, same era.

Then

Learning graphical model parameters with approximate marginal inference
Domke, Justin · 2013
Later among the works it cites.
Austerity in mcmc land: Cutting the metropolis-hastings budget
Korattikara, Anoop, Chen, Yutian, and Welling, Max · 2014
Closest in time.
Neural variational inference and learning in belief networks
Original
Mnih, Andriy and Gregor, Karol · 2014
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…