I tried to run testsmf.ssc on Netflix dataset. For all the cases d=32 to d=512 RMSE is 1.05, which cannot be true. I tried running testsfa.ssc script on the same dataset. There I get RMSE ranging from .92 to .95 which is acceptable. Is the prediction function wrong? I tried to use same prediction function but that did not work either.
Any kind of help will be really appreciated.
I tried to run testsmf.ssc on Netflix dataset. For all the cases d=32 to d=512 RMSE is 1.05, which cannot be true. I tried running testsfa.ssc script on the same dataset. There I get RMSE ranging from .92 to .95 which is acceptable. Is the prediction function wrong? I tried to use same prediction function but that did not work either.
Any kind of help will be really appreciated.