We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
clarify gd_mf_weights
fixed path to build directory, clarified -i model.reg in gd_mf_weights
Updated Matrix factorization example (markdown)
show an example of the matrix factorization training data from the movie dataset
typo
no need to specify -q and --rank in testing mode. They seem to be read from the model
add /dev/stdin when vw gets its input from a pipe (otherwise doesn't work for me with latest vw)
update the movielens data URL (location has changed), packing method and directory
Another regularization -> l2 fix, plus fix one other misspelling
Update --regularization to the newer --l2 (l2 norm lambda) option
Created Matrix factorization example (markdown)