zzmdl

07-17-2008, 11:54 AM

Hi,

When I choose bayes net, no matter what kind of structure learning method I use, the structure is the same as naive bayes. That is only one root of the tree, which is the class label, and all other attributes are the children of the root. Is there anyone also face this problem before? Is there any thing I missed? Thanks a lot!

Best regards

zzmdl

When I choose bayes net, no matter what kind of structure learning method I use, the structure is the same as naive bayes. That is only one root of the tree, which is the class label, and all other attributes are the children of the root. Is there anyone also face this problem before? Is there any thing I missed? Thanks a lot!

Best regards

zzmdl