I'm trying to use the MatrixFactorizationModel to, for instance, determine the latent factors of a user or item that were not used in the training data of the model. I'm not as concerned about the rating as I am with the latent factors for the user/item.
Thanks!
