|
|
| Line 1: |
Line 1: |
| In [[machine learning]], '''PU learning''' is a collection of [[Semi-supervised learning|semisupervised]] techniques for training [[Statistical classification|binary classifiers]] on '''p'''ositive and '''u'''nlabeled examples only.<ref>{{cite book
| | Частное предприятие «Илигран»<br>220073, г. [http://iligran.by/%d0%bd%d0%b0%d1%88%d0%b0-%d1%82%d0%b5%d1%85%d0%bd%d0%b8%d0%ba%d0%b0-%d0%b2-%d0%b0%d1%80%d0%b5%d0%bd%d0%b4%d1%83/ аренда башенных кранов Минск], ул. Кальварийская, дом 25, офис 424<br>Телефоны:<br><br>+375 44 545-67-00<br><br>+375 29 379-91-88<br>+375 17 204 42 28 (факс)<br>+375 17 204 42 26 (факс)<br>+375 17 204 01 72<br>Email: 2044228@mail.ru<br><br>http://iligran.by |
| |last=Liu
| |
| |first=Bing
| |
| |title=Web Data Mining
| |
| |publisher=Springer
| |
| |year=2007
| |
| |pages=165−178
| |
| }}</ref>
| |
| | |
| In PU learning, two sets of samples are assumed to be available for training: the positive set <math>P</math> and a ''mixed set'' <math>U</math>, which is assumed to contain both positive and negative samples, but without these being labeled as such. This contrasts with other forms of semisupervised learning, where it is assumed that a labeled set containing examples of both classes is available. A variety of techniques exist to adapt [[supervised learning|supervised]] classifiers to the PU learning setting. PU learning successfully been applied to [[text classification]] <ref>{{cite conference |authors=Bing Liu, Wee Sun Lee, [[Philip S. Yu]] and Xiao-Li Li |year=2002 |title=Partially supervised classification of text documents |conference=ICML |pages=8–12}}</ref><ref>{{cite conference |authors=Hwanjo Yu, Jiawei Han, Kevin Chen-Chuan Chang |title=PEBL: positive example based learning for web page classification using SVM |conference=ACM SIGKDD |year=2002}}</ref><ref>{{cite conference |authors=Xiao-Li Li and Bing Liu |title=Learning to classify text using positive and unlabeled data |conference=IJCAI |year=2003}}</ref> and [[Bioinformatics]] tasks.<ref>{{cite conference |authors=Peng Yang, Xiao-Li Li, Jian-Ping Mei, Chee-Keong Kwoh and See-Kiong Ng |title=Positive-Unlabeled Learning for Disease Gene Identification |conference=Bioinformatics, Vol 28(20)|year=2012}}</ref>
| |
| | |
| ==References==
| |
| <references/> | |
| | |
| [[Category:Machine learning]]
| |
| | |
| | |
| {{compu-sci-stub}}
| |
Latest revision as of 13:50, 23 April 2014
Частное предприятие «Илигран»
220073, г. аренда башенных кранов Минск, ул. Кальварийская, дом 25, офис 424
Телефоны:
+375 44 545-67-00
+375 29 379-91-88
+375 17 204 42 28 (факс)
+375 17 204 42 26 (факс)
+375 17 204 01 72
Email: 2044228@mail.ru
http://iligran.by