Variational Message Passing¶
Winn, J. M., & Bishop, C. M. (2005). Variational Message Passing. Journal of Machine Learning Research.
Cited by¶
1 citation across 1 artifact.
Each citation links to the sentence it supports in the citing article.
Domain-specific¶
- Variational Message Passing
- It is the compilation of that identity into local messages for conjugate-exponential Bayesian networks: parent-to-child messages carry expected sufficient statistics, child-to-parent messages carry contributions to natural parameters, and a node's updated variational distribution is formed from the incoming contributions
This sourceThe paper names exactly this class - hidden Markov models, Kalman filters, factor analysers, principal component analysers, and mixtures and hierarchical mixtures of these - as special cases of its conjugate-exponential architecture (section 4 sets out the allowable building blocks); the Gaussian mixture is the case worked through in full.
- It is the compilation of that identity into local messages for conjugate-exponential Bayesian networks: parent-to-child messages carry expected sufficient statistics, child-to-parent messages carry contributions to natural parameters, and a node's updated variational distribution is formed from the incoming contributions
Verification¶
This reference passed the adversarial substantiation pipeline: it was checked to exist and to support the claim it is attached to. See how references were verified.
Registry ID ref:335230a557a4 · see in the full table