As the Netflix Prize competition has demonstrated, matrix factorization models are superior to classic nearest-neighbor techniques for producing product recommendations, allowing the incorporation of additional information such as implicit feedback, temporal effects, and confidence levels.
Veranschaulicht deutlich, den Lernprozess eines Maschinenlern-Algorithmus, auch wenn nicht genau zum Thema passend.
References
Bookmarks
deleting review
Please log in to take part in the discussion (add own reviews or comments).
Cite this publication
More citation styles
- please select -
%0 Journal Article
%1 Koren:2009:MFT:1608565.1608614
%A Koren, Yehuda
%A Bell, Robert
%A Volinsky, Chris
%C Los Alamitos, CA, USA
%D 2009
%I IEEE Computer Society Press
%J Computer
%K
%N 8
%P 30--37
%R 10.1109/MC.2009.263
%T Matrix Factorization Techniques for Recommender Systems
%U http://dx.doi.org/10.1109/MC.2009.263
%V 42
%X As the Netflix Prize competition has demonstrated, matrix factorization models are superior to classic nearest-neighbor techniques for producing product recommendations, allowing the incorporation of additional information such as implicit feedback, temporal effects, and confidence levels.
@article{Koren:2009:MFT:1608565.1608614,
abstract = {As the Netflix Prize competition has demonstrated, matrix factorization models are superior to classic nearest-neighbor techniques for producing product recommendations, allowing the incorporation of additional information such as implicit feedback, temporal effects, and confidence levels.},
acmid = {1608614},
added-at = {2013-11-08T23:19:52.000+0100},
address = {Los Alamitos, CA, USA},
author = {Koren, Yehuda and Bell, Robert and Volinsky, Chris},
biburl = {https://www.bibsonomy.org/bibtex/26375c7802c40dd797c7b043f873b8a8f/florian.pf},
doi = {10.1109/MC.2009.263},
interhash = {cface72aeba6ee8c561ccd15035d0ead},
intrahash = {6375c7802c40dd797c7b043f873b8a8f},
issn = {0018-9162},
issue_date = {August 2009},
journal = {Computer},
keywords = {},
month = aug,
number = 8,
numpages = {8},
pages = {30--37},
publisher = {IEEE Computer Society Press},
timestamp = {2013-11-08T23:19:52.000+0100},
title = {Matrix Factorization Techniques for Recommender Systems},
url = {http://dx.doi.org/10.1109/MC.2009.263},
volume = 42,
year = 2009
}