k-fold cross validation

kck · Post by **kck** » Sat Dec 06, 2008 3:06 am

is k-fold cross validation possible for bayesian networks including own crafted one using smile or genie?

mark · Post by **mark** » Mon Dec 08, 2008 4:31 pm

Depends on what you're trying to evaluate. Please elaborate.

kck · Post by **kck** » Wed Dec 10, 2008 2:21 am

To evaluate the prediction accuracy of a BN given a test dataset.
For instance, 10-fold cross validation.
Break data into 10 sets of size n/10. Train on 9 datasets and test on 1. Repeat 10 times and take a mean accuracy.

mark · Post by **mark** » Wed Dec 10, 2008 2:44 am

Yes, that's certainly possible but you'll have to write it yourself.