Semi-supervised Regression with Order Preferences
File(s)
Date
2006Author
Zhu, Xiaojin
Goldberg, Andrew
Publisher
University of Wisconsin-Madison Department of Computer Sciences
Metadata
Show full item recordAbstract
Following a discussion on the general form of regularization for semi-supervised learning, we propose a semi-supervised regression algorithm. It is based on the assumption that we have certain order preferences on unlabeled data (e.g., point X1 has a larger target value than x2). Semi-supervised learning consists of enforcing the order preferences as regularization in a risk minimization framework. The optimization problem can be effectivley solved by a linear program. Experiments show that the proposed semi-supervised regression outperforms standrad regression.
Permanent Link
http://digital.library.wisc.edu/1793/60530Citation
TR1578