## Learning from Labeled and Unlabeled Data using Graph Mincuts (2001)

### Cached

### Download Links

- [www.aladdin.cs.cmu.edu]
- [www.cs.cmu.edu]
- DBLP

### Other Repositories/Bibliography

Citations: | 281 - 6 self |

### BibTeX

@MISC{Blum01learningfrom,

author = {Avrim Blum and Shuchi Chawla},

title = {Learning from Labeled and Unlabeled Data using Graph Mincuts},

year = {2001}

}

### Years of Citing Articles

### OpenURL

### Abstract

Many application domains suffer from not having enough labeled training data for learning. However, large amounts of unlabeled examples can often be gathered cheaply. As a result, there has been a great deal of work in recent years on how unlabeled data can be used to aid classification. We consider an algorithm based on finding minimum cuts in graphs, that uses pairwise relationships among the examples in order to learn from both labeled and unlabeled data. Our algorithm

### Citations

8936 | Cache-oblivious algorithms - Frigo, Leiserson, et al. - 1999 |

2750 | Normalized cuts and image segmentation - Shi, Malik - 1997 |

1773 | A global geometric framework for nonlinear dimensionality reduction - TENENBAUM, SILVA, et al. - 2000 |

1707 | Nonlinear Dimensionality Reduction by Locally Linear Embedding". Science December 2000 - Roweis, Saul |

1306 | Combining labeled and unlabeled data with co-training - Blum, Mitchell - 1998 |

573 | Inducing features of random fields - Pietra, Stephen, et al. - 1997 |

342 | Exact maximum a posteriori estimation for binary images - Greig, Porteous, et al. - 1989 |

286 | An optimal graph theoretic approach to data clustering: theory and its application to image segmentation - Wu, Leahy - 1993 |

239 | A maximum-flow formulation of the n-camera stereo correspondence problem
- Roy, Cox
- 1998
(Show Context)
Citation Context ... Zhang & Oles, 2000; Schuurmans, 1997). Recently, a method based on graph mincuts has been proposed in the vision literature for the problem of cleaning up 3-D pixel images (Greig et al., 1989; Roy & =-=Cox, 1998; Boy-=-kov et al., 1998; Snow et al., 2000). Given an initial noisy image created from a stereo camera, the goal is to improve the image by minimizing an appropriate \energy function." This energy funct... |

190 | Semi-supervised support vector machines - Bennett, Demiriz - 1999 |

175 | Markov Random Fields with efficient approximations - Boykov, Veksler, et al. - 1998 |

171 | Learning to classify text from labeled and unlabeled documents
- Nigam, McCallum, et al.
- 1998
(Show Context)
Citation Context ... a result, there has been a good deal of work in recent years on how unlabeled data can be usefully employed in order to produce better predictions (Ratsaby & Venkatesh, 1995; Castelli & Cover, 1996; =-=Nigam et al., 1998-=-; Blum & Mitchell, 1998; Bennett & Demiriz, 1998; Hofmann, 1999; Zhang & Oles, 2000; Schuurmans, 1997). Recently, a method based on graph mincuts has been proposed in the vision literature for the pro... |

164 | Approximation algorithms for classification problems with pairwise relationships: Metric labeling and markov random fields - Kleinberg, Tardos - 1999 |

102 | The relative value of labeled and unlabeled samples in pattern recognition with an unknown mixing parameter - Castelli, Cover - 1996 |

86 | Exact Voxel Occupancy with Graph Cuts - Snow, Viola, et al. - 2000 |

43 | A new metric-based approach to model selection
- Schuurmans
- 1997
(Show Context)
Citation Context ...loyed in order to produce better predictions (Ratsaby & Venkatesh, 1995; Castelli & Cover, 1996; Nigam et al., 1998; Blum & Mitchell, 1998; Bennett & Demiriz, 1998; Hofmann, 1999; Zhang & Oles, 2000; =-=Schuurmans, 1997-=-). Recently, a method based on graph mincuts has been proposed in the vision literature for the problem of cleaning up 3-D pixel images (Greig et al., 1989; Roy & Cox, 1998; Boykov et al., 1998; Snow ... |

37 |
Learning from a mixture of labeled and unlabeled examples with parametric side information
- Ratsaby, Venkatesh
- 1995
(Show Context)
Citation Context ... of which have been classied by hand. As a result, there has been a good deal of work in recent years on how unlabeled data can be usefully employed in order to produce better predictions (Ratsaby & V=-=enkatesh, 1995-=-; Castelli & Cover, 1996; Nigam et al., 1998; Blum & Mitchell, 1998; Bennett & Demiriz, 1998; Hofmann, 1999; Zhang & Oles, 2000; Schuurmans, 1997). Recently, a method based on graph mincuts has been p... |

22 | Inducing features of random - Pietra, S, et al. - 1997 |

13 | Markov random with ecient approximations - Boykov, Veksler, et al. - 1998 |

10 | Approximation Algorithms for Classi Problems with Pairwise Relationships: Metric Labeling and Markov Random Fields - Kleinberg - 1999 |

8 |
Text categorization with labeled and unlabeled data: A generative model approach
- Hofmann
- 1999
(Show Context)
Citation Context ... unlabeled data can be usefully employed in order to produce better predictions (Ratsaby & Venkatesh, 1995; Castelli & Cover, 1996; Nigam et al., 1998; Blum & Mitchell, 1998; Bennett & Demiriz, 1998; =-=Hofmann, 1999-=-; Zhang & Oles, 2000; Schuurmans, 1997). Recently, a method based on graph mincuts has been proposed in the vision literature for the problem of cleaning up 3-D pixel images (Greig et al., 1989; Roy &... |

6 |
Repository of machine learning databases. http://www.ics.uci.edu/ mlearn/MLRepository.html
- UCI
- 2000
(Show Context)
Citation Context ...ction 4. 5.1 Standard datasets We compared the mincut algorithm with two standard learning algorithms, ID3 and 3-nearest-neighbor, on datasets obtained from the UC Irvine Machine Learning Repository (=-=UCI, 20-=-00). The mincut algorithm has many degrees of freedom in terms of how the edge weights are dened. In order to make the experiments as clean as possible, we consider weighting functions specically moti... |

1 | Text categorization with labeled and unlabeled data: A generatire model approach - Holmann - 1999 |