Hi,

does anyone know of a principled approach of choosing the number of features for ALS (other than cross-validation?)

--sebastian

Reply via email to