Is it possible to do evaluation with mymedialite tool ‘’rating_prediction’’ using splitted dataset (movielens 10m) for five-fold cross-validation of rating predictions. Dataset is splitted in r1.test.... r5.test, r1.train.... r5.train. I have managed to execute evaluation with full dataset (ratings.dat) with command : „rating_prediction --recommender=BiasedMatrixFactorization --file-format=movielens_1m --training-file=ratings.dat --data-dir=ml-10m --cross-validation=5“, but have no idea how to provide it with splitted dataset.
Is it possible to do evaluation with mymedialite tool ‘’rating_prediction’’ using splitted dataset (movielens 10m) for five-fold cross-validation of rating predictions. Dataset is splitted in r1.test.... r5.test, r1.train.... r5.train. I have managed to execute evaluation with full dataset (ratings.dat) with command : „rating_prediction --recommender=BiasedMatrixFactorization --file-format=movielens_1m --training-file=ratings.dat --data-dir=ml-10m --cross-validation=5“, but have no idea how to provide it with splitted dataset.
--
You received this message because you are subscribed to the Google Groups "MyMediaLite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mymedialite...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
You received this message because you are subscribed to a topic in the Google Groups "MyMediaLite" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mymedialite/O6JbBS0mv6k/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mymedialite...@googlegroups.com.