AldenBlack
Suppose I learn a network from data, and then wish to manually add utility and decision nodes to the network to assist with decision making. When I try to do so, Bayesialab tells me, obviously, that doing so will remove the association between the database and the network since the utility and decision nodes don't have any representation in the data set.What analysis capabilities will I lose if I remove the association with the dataset? Obviously I won't be able to perform any more learning in modeling mode, but will I lose any capabilities or accuracy when performing tasks in validation mode?
Quote 0 0
Dan
Indeed you will no longer be able to use the learning algorithms, nor use the functions like Batch Inference that require a dataset. However, as analysis is concerned, you should not loose any feature.
Quote 0 0
Joe_Black
After you add the new nodes manually you can re-associate the dataset ensuring that you have new columns in the dataset for the manually added nodes if need be.
Quote 0 0