Aleks Jakulin and Irina Rish (2006) Bayesian Learning of Markov Network Structure. In: ECML 2006, August 2006, Berlin, Germany.
This is the latest version of this item.
We propose a simple and efficient approach to building undirected probabilistic classification models (Markov networks) that extend naive Bayes classifiers and outperform existing directed probabilistic classifiers (Bayesian networks) of similar complexity. Our Markov network model is represented as a set of consistent probability distributions on subsets of variables. Inference with such a model can be done efficiently in closed form for problems like class probability estimation. We also propose a highly efficient Bayesian structure learning algorithm for conditional prediction problems, based on integrating along a hill-climb in the structure space. Our prior based on the degrees of freedom effectively prevents overfitting.
|Item Type: ||Conference or Workshop Item (Paper)|
|Keywords: ||Kikuchi approximation, Bayesian statistics, Bayesian networks|
|Language of Content: ||English|
|Related URLs: |
|Institution: ||Columbia University|
|Department: ||Department of Statistics|
|Divisions: ||Faculty of Computer and Information Science > Other|
|Item ID: ||229|
|Date Deposited: ||10 Jan 2007|
|Last Modified: ||02 Dec 2013 11:49|
Available Versions of this Item.
Actions (login required)