Semisupervised classification with cluster regularization

Rodrigo G. F. SOARES, Huanhuan CHEN, Xin YAO

Research output: Journal PublicationsJournal Article (refereed)peer-review

52 Citations (Scopus)

Abstract

Semisupervised classification (SSC) learns, from cheap unlabeled data and labeled data, to predict the labels of test instances. In order to make use of the information from unlabeled data, there should be an assumed relationship between the true class structure and the data distribution. One assumption is that data points clustered together are likely to have the same class label. In this paper, we propose a new algorithm, namely, cluster-based regularization (ClusterReg) for SSC, that takes the partition given by a clustering algorithm as a regularization term in the loss function of an SSC classifier. ClusterReg makes predictions according to the cluster structure together with limited labeled data. The experiments confirmed that ClusterReg has a good generalization ability for real-world problems. Its performance is excellent when data follows this cluster assumption. Even when these clusters have misleading overlaps, it still outperforms other state-of-the-art algorithms. © 2012 IEEE.
Original languageEnglish
Article number6317193
Pages (from-to)1779-1792
Number of pages14
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume23
Issue number11
Early online date1 Oct 2012
DOIs
Publication statusPublished - Nov 2012
Externally publishedYes

Bibliographical note

This work was supported in part by The Capes Foundation, Ministry of Education of Brazil, Brazil, and the European Union Seventh Framework Programme under Grant 270428.; Funding text 2: Mr. Soares is a recipient of a scholarship from the Capes Foundation, Brazil, and the Brazilian Council for Scientific and Technological Development Scholarship.

Keywords

  • Clustering
  • machine learning
  • regularization
  • semisupervised learning

Fingerprint

Dive into the research topics of 'Semisupervised classification with cluster regularization'. Together they form a unique fingerprint.

Cite this