Transductive Learning via Spectral Graph Partitioning
T. Joachims. Proceedings of the Twentieth International Conference on Machine Learning (ICML-2003), (2003)
Abstract
We present a new method for transductive learning, which can be seen as a transductive version of the k nearest-neighbor classifier. Unlike for many other transductive learning methods, the training problem has a meaningful relaxation that can be solved globally optimally using spectral methods. We propose an algorithm that robustly achieves good generalization performance and that can be trained efficiently. A key advantage of the algorithm is that it does not require additional heuristics to avoid unbalanced splits. Furthermore, we show a connection to transductive Support Vector Machines, and that an effective Co-Training algorithm arises as a special case.
%0 Conference Paper
%1 citeulike:105008
%A Joachims, Thorsten
%B Proceedings of the Twentieth International Conference on Machine Learning (ICML-2003)
%D 2003
%K clustering contraint spectral
%T Transductive Learning via Spectral Graph Partitioning
%X We present a new method for transductive learning, which can be seen as a transductive version of the k nearest-neighbor classifier. Unlike for many other transductive learning methods, the training problem has a meaningful relaxation that can be solved globally optimally using spectral methods. We propose an algorithm that robustly achieves good generalization performance and that can be trained efficiently. A key advantage of the algorithm is that it does not require additional heuristics to avoid unbalanced splits. Furthermore, we show a connection to transductive Support Vector Machines, and that an effective Co-Training algorithm arises as a special case.
@inproceedings{citeulike:105008,
abstract = {We present a new method for transductive learning, which can be seen as a transductive version of the k nearest-neighbor classifier. Unlike for many other transductive learning methods, the training problem has a meaningful relaxation that can be solved globally optimally using spectral methods. We propose an algorithm that robustly achieves good generalization performance and that can be trained efficiently. A key advantage of the algorithm is that it does not require additional heuristics to avoid unbalanced splits. Furthermore, we show a connection to transductive Support Vector Machines, and that an effective Co-Training algorithm arises as a special case.},
added-at = {2007-04-28T16:01:00.000+0200},
author = {Joachims, Thorsten},
biburl = {https://www.bibsonomy.org/bibtex/227daef0514080660b9e25e27a5d41b64/kzhou},
booktitle = {Proceedings of the Twentieth International Conference on Machine Learning (ICML-2003)},
citeulike-article-id = {105008},
interhash = {1686cf7cd6cc2b7b0825148c4742b8da},
intrahash = {27daef0514080660b9e25e27a5d41b64},
keywords = {clustering contraint spectral},
priority = {4},
timestamp = {2007-04-28T16:01:00.000+0200},
title = {Transductive Learning via Spectral Graph Partitioning},
year = 2003
}