By using CI (conditional independence) tests, it can be pruned a fully connected potential graph to a best PG, which is expected to approximate the undirected version of the underlying directed graph.

It shows that the distribution determined by the Bayesian network maximises entropy given the causal and probability distribution of a Bayesian network under the conditional independence.

By giving distinct classification rule to each sample,decision forest rather than decision tree is constructed to automatically determine relatively independent attribute subset,and based on this we integrate submodel by applying conditional independence assumption.

Conditional regressive independence and conditional independence;