SVDFeature is a machine learning toolkit for feature-based collaborative filtering. SVDFeature is designed to efficiently solve the feature-based matrix factorization. The feature-based setting allows us to build factorization models incorporating side information such as temporal dynamics, neighborhood relationship, and hierarchical information. The toolkit is capable of both rate prediction and collaborative ranking, and is carefully designed for efficient training on large-scale data set. Using this toolkit, we built solutions to win KDD Cup for two consecutive years.
- Changes to previous version:
JMLR MLOSS version.
Other available revisons
Version Changelog Date 1.2.2
JMLR MLOSS version.
January 9, 2013, 02:21:18 1.2
Change the code path structure to allow easier inclusion of extensions. Start a new project homepage. Add the most recent version of extensions(though some are not yet documented) code used during KDDCup 2012.
July 24, 2012, 10:32:10 1.1.6
Add lite version of SVDFeature, used as example code for simple SGD implementation of the algorithm.
December 2, 2011, 12:14:48 1.1.5
Add support for common parameter space of user/item features. This allows interesting usage such as collective(joint) matrix factorization and symmetric relation prediction.
November 21, 2011, 08:21:52 220.127.116.11
No change in the toolkit, add a detailed non-trivial collaborative ranking experiment KDDCup'11 track2 using SVDFeature, getting state-of-art performance. The link to the experiment scripts are in our project page
November 2, 2011, 09:27:17 1.1.4
fix a bug happened in rank-model. More robust input buffering for user grouped format, less memory cost for loading input data when doing SVD++ and rank training.
October 12, 2011, 02:37:27 1.1.3
Add detailed regularization settings, fix a bug that only occurs in windows
October 7, 2011, 10:07:08 1.1.2
Add more regularization options
September 26, 2011, 05:00:47 1.1.1
The project is now self-contained. Add document of major interfaces in the project.
September 19, 2011, 12:30:07 1.1
Initial Announcement on mloss.org.
September 18, 2011, 07:18:39
No one has posted any comments yet. Perhaps you'd like to be the first?
Leave a comment
You must be logged in to post comments.