Consistent estimation of the basic neighborhood of Markov random fields
By Imre Csiszár, Zsolt Talata. 2004
78 citations recorded by the source corpus.
For Markov random fields with finite set of states, a modification of the Bayesian information criterion admits strongly consistent estimation of the smallest region that determines the conditional distributions. Phase transition or nonstationarity do not affect the result.
Subfields: Artificial Intelligence, Mathematical Physics
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
In Mathematical Physics
Citations in the corpus
- Learning generative models for protein fold families (2011, neutral): “While it may be expected that using the pseudo log-likelihood instead of the true log-likelihood may in fact lead to inconsistent selection a somewhat surprising result [43] shows that in the case of BIC using pseudo log-likelihood is in fact also consistent for model selection.”
- Finding and Leveraging Structure in Learning Problems (2012, result): “While it may be expected that using the pseudo log-likelihood instead of the true log-likelihood may in fact lead to inconsistent selection a somewhat surprising result of Csiszar and Talata [51] shows that in the case of BIC using pseudo log-likelihood is in fact also consistent for model selection”