What reasons exist to favor bayesian networks over other, presumably more powerful, predicitve models like random forests or deep neural networks?
$\begingroup$
$\endgroup$
1
-
2$\begingroup$ i am less familiar with DNN's - but i would presume that one of the reasons to continue to use Bayesian networks is that they are probabilistic models, that more intuitively (at least than random forests) give a probabilistic interpretation of outcomes and results $\endgroup$– WouterCommented Jul 20, 2015 at 13:44
Add a comment
|