Jump to ContentJump to Main Navigation
Semi-Supervised Learning - MIT Press Scholarship Online
Users without a subscription are not able to see the full content.

Semi-Supervised Learning

Olivier Chapelle, Bernhard Scholkopf, and Alexander Zien


In the field of machine learning, semi-supervised learning (SSL) occupies the middle ground, between supervised learning (in which all training examples are labeled) and unsupervised learning (in which no label data are given). Interest in SSL has increased in recent years, particularly because of application domains in which unlabeled data are plentiful, such as images, text, and bioinformatics. This overview of SSL presents state-of-the-art algorithms, a taxonomy of the field, selected applications, benchmark experiments, and perspectives on ongoing and future research. It first presents the ... More

Keywords: machine learning, semi-supervised learning, training examples, label data, application domains, bioinformatics, algorithms, taxonomy, low-density separation, manifold structure

Bibliographic Information

Print publication date: 2006 Print ISBN-13: 9780262033589
Published to MIT Press Scholarship Online: August 2013 DOI:10.7551/mitpress/9780262033589.001.0001


Affiliations are at time of print publication.

Olivier Chapelle, editor

Bernhard Scholkopf, editor

Alexander Zien, editor

Show Summary Details

subscribe or login to access all content.



1 Introduction to Semi-Supervised Learning

Chapelle Olivier, Schölkopf Bernhard, and Zien Alexander

I Generative Models

1 A Taxonomy for Semi-Supervised Learning Methods


3 Semi-Supervised Text Classification Using EM

knigam@kamalnigam.com mccallum@cs.umass.edu tom.mitchell@cmu.edu

5 Probabilistic Semi-Supervised Clustering with Constraints

mbilenko@cs.utexas.edu sugato@cs.utexas.edu abanerje@ece.utexas.edu mooney@cs.utexas.edu

II Low-Density Separation

7 Semi-Supervised Learning Using Semi-Definite Programming

tijl.debie@gmail.com nello@support-vector.net

8 Gaussian Processes and the Null-Category Noise Model

neil@dcs.shef.ac.uk jordan@cs.berkeley.edu

9 Entropy Regularization

yves.grandvalet@utc.fr bengioy@iro.umontreal.ca

10 Data-Dependent Regularization

adrianc@mit.edu tommi@csail.mit.edu

III Graph-Based Methods

11 Label Propagation and Quadratic Criterion

bengioy@iro.umontreal.ca delallea@iro.umontreal.ca nicolas.le.roux@umontreal.ca

12 The Geometric Basis of Semi-Supervised Learning

vikass@cs.uchicago.edu mbelkin@cse.ohio-state.edu niyogi@cs.uchicago.edu

13 Discrete Regularization

dengyong.zhou@tuebingen.mpg.de bernhard.schoelkopf@tuebingen.mpg.de

14 Semi-Supervised Learning with Conditional Harmonic Mixing

Chris.Burges@microsoft.com jplatt@microsoft.com

IV Change of Representation

15 Graph Kernels by Spectral Transforms

zhuxj@cs.cmu.edu jkandola@gatsby.ucl.ac.uk lafferty@cs.cmu.edu zoubin@eng.cam.ac.uk

16 Spectral Methods for Dimensionality Reduction

lsaul@cis.upenn.edu kilianw@seas.upenn.edu feisha@cis.upenn.edu jhham@seas.upenn.edu ddlee@seas.upenn.edu

17 Modifying Distances

Sajama ssajama@ieng9.ucsd.edu alqn@ucsd.edu

V Semi-Supervised Learning in Practice

18 Large-Scale Algorithms

delallea@iro.umontreal.ca bengioy@iro.umontreal.ca nicolas.le.roux@umontreal.ca

19 Semi-Supervised Protein Classification Using Cluster Kernels

jasonw@nec-labs.com cleslie@cs.columbia.edu eie@cs.columbia.edu noble@gs.washington.edu

20 Prediction of Protein Function from Networks

shin@tuebingen.mpg.de koji.tsuda@tuebingen.mpg.de

25 Analysis of Benchmarks

Chapelle Olivier, Schölkopf Bernhard, and Zien Alexander

VI Perspectives

22 An Augmented PAC Model for Semi-Supervised Learning

ninamf@cs.cmu.edu avrim@cs.cmu.edu

23 Metric-Based Approaches for Semi-Supervised Regression and Classification

dale@cs.ualberta.ca finnwork@lucubratio.org d3wilkinson@cs.uwaterloo.ca yuhong@cs.ualberta.ca

25 A Discussion of Semi-Supervised Learning and Transduction

Chapelle Olivier, Schölkopf Bernhard, and Zien Alexander