Main Page | Modules | Data Structures | File List | Globals | Related Pages

bnlearn File Reference

Detailed Description

Learns the structure of a BeliefNet from a data set. Designed to be easily modified.

Learns the structure and parameters of a Bayesian network using the standard method. The structural prior is set by using P(S) = kappa ^ (symetric difference between the net and the prior net). The parameter prior is K2. This is basically a wrapper around the code in bnlearn-engine.h. This program is a stripped down version of the vfbn learner, done so that the code would be clearer and easier to modify.

bnlearn can load training data into RAM, if space is available, or it can scan data repeatedly from disk.

See the documentation for the bnlearn-engine for even more information about the parameters and their meanings.

bnlearn takes input and does output in c4.5 format. It expects to find the files <stem>.names and <stem>.data.

to Matthew Richardson for making substantial optimizations to the bnlearn program.

Wish List:
A version of this program that is intelligent about dealing with unobserved (or partially observed) variables.


Generated for VFML by doxygen hosted by Logo