There is not necessarily a relation, but, a "good" recommender ought to be good at predicting ratings, and ought to return good recommendations. So yes you would generally expect a low error when you get a high precision, but there is not a direct connection.
On Wed, Mar 28, 2012 at 5:52 PM, ziad kamel <ziad.kame...@gmail.com> wrote: > In evaluation , is there a relation between score and precision ? > double score = evaluator.evaluate(recommenderBuilder, null, model, 0.95, > 0.5); > > For example, when we get low score will that mean we will get high > precision ? >
