Kikuchi-Bayes: Factorized Models for Approximate Classification in Closed Form

We propose a simple family of classification models, based on the Kikuchi approximation to free energy. We note that the resulting product of potentials is not normalized, but for classification it is easy to perform the normalization for each instance separately. We propose a learning method based on including those initial regions that would otherwise be significantly different from those estimated directly. We observe that this algorithm outperforms other methods, such as the tree-augmented naive Bayes, but that the inclusion of regions may increase the approximation error, even in cases when adding a region does not yield loopy dependencies.