Joint sparse regularization based Sparse Semi-Supervised Extreme Learning Machine (S3ELM) for classification

Xiaozhuo Luo, F. Liu, Shuyuan Yang, Xiaodong Wang, Zhiguo Zhou

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Extreme Learning Machine (ELM) has received increasing attention for its simple principle, low computational cost and excellent performance. However, a large number of labeled instances are often required, and the number of hidden nodes should be manually tuned, for better learning and generalization of ELM. In this paper, we propose a Sparse Semi-Supervised Extreme Learning Machine (S3ELM) via joint sparse regularization for classification, which can automatically prune the model structure via joint sparse regularization technology, to achieve more accurate, efficient and robust classification, when only a small number of labeled training samples are available. Different with most of greedy-algorithms based model selection approaches, by using ℓ2,1-norm, S3ELM casts a joint sparse constraints on the training model of ELM and formulate a convex programming. Moreover, with a Laplacian, S3ELM can make full use of the information from both the labeled and unlabeled samples. Some experiments are taken on several benchmark datasets, and the results show that S3ELM is computationally attractive and outperforms its counterparts.

Original languageEnglish (US)
Pages (from-to)149-160
Number of pages12
JournalKnowledge-Based Systems
Volume73
DOIs
StatePublished - Jan 1 2015

Fingerprint

Learning systems
Convex optimization
Model structures
Machine learning
Regularization
Costs
Experiments

Keywords

  • Extreme learning machine
  • Joint sparse regularization
  • Laplacian
  • Sparse semi-supervised learning
  • ℓ -Norm

ASJC Scopus subject areas

  • Software
  • Management Information Systems
  • Information Systems and Management
  • Artificial Intelligence

Cite this

Joint sparse regularization based Sparse Semi-Supervised Extreme Learning Machine (S3ELM) for classification. / Luo, Xiaozhuo; Liu, F.; Yang, Shuyuan; Wang, Xiaodong; Zhou, Zhiguo.

In: Knowledge-Based Systems, Vol. 73, 01.01.2015, p. 149-160.

Research output: Contribution to journalArticle

@article{6a2436aed62c4c399466f9b346cf7d44,
title = "Joint sparse regularization based Sparse Semi-Supervised Extreme Learning Machine (S3ELM) for classification",
abstract = "Extreme Learning Machine (ELM) has received increasing attention for its simple principle, low computational cost and excellent performance. However, a large number of labeled instances are often required, and the number of hidden nodes should be manually tuned, for better learning and generalization of ELM. In this paper, we propose a Sparse Semi-Supervised Extreme Learning Machine (S3ELM) via joint sparse regularization for classification, which can automatically prune the model structure via joint sparse regularization technology, to achieve more accurate, efficient and robust classification, when only a small number of labeled training samples are available. Different with most of greedy-algorithms based model selection approaches, by using ℓ2,1-norm, S3ELM casts a joint sparse constraints on the training model of ELM and formulate a convex programming. Moreover, with a Laplacian, S3ELM can make full use of the information from both the labeled and unlabeled samples. Some experiments are taken on several benchmark datasets, and the results show that S3ELM is computationally attractive and outperforms its counterparts.",
keywords = "Extreme learning machine, Joint sparse regularization, Laplacian, Sparse semi-supervised learning, ℓ -Norm",
author = "Xiaozhuo Luo and F. Liu and Shuyuan Yang and Xiaodong Wang and Zhiguo Zhou",
year = "2015",
month = "1",
day = "1",
doi = "10.1016/j.knosys.2014.09.014",
language = "English (US)",
volume = "73",
pages = "149--160",
journal = "Knowledge-Based Systems",
issn = "0950-7051",
publisher = "Elsevier",

}

TY - JOUR

T1 - Joint sparse regularization based Sparse Semi-Supervised Extreme Learning Machine (S3ELM) for classification

AU - Luo, Xiaozhuo

AU - Liu, F.

AU - Yang, Shuyuan

AU - Wang, Xiaodong

AU - Zhou, Zhiguo

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Extreme Learning Machine (ELM) has received increasing attention for its simple principle, low computational cost and excellent performance. However, a large number of labeled instances are often required, and the number of hidden nodes should be manually tuned, for better learning and generalization of ELM. In this paper, we propose a Sparse Semi-Supervised Extreme Learning Machine (S3ELM) via joint sparse regularization for classification, which can automatically prune the model structure via joint sparse regularization technology, to achieve more accurate, efficient and robust classification, when only a small number of labeled training samples are available. Different with most of greedy-algorithms based model selection approaches, by using ℓ2,1-norm, S3ELM casts a joint sparse constraints on the training model of ELM and formulate a convex programming. Moreover, with a Laplacian, S3ELM can make full use of the information from both the labeled and unlabeled samples. Some experiments are taken on several benchmark datasets, and the results show that S3ELM is computationally attractive and outperforms its counterparts.

AB - Extreme Learning Machine (ELM) has received increasing attention for its simple principle, low computational cost and excellent performance. However, a large number of labeled instances are often required, and the number of hidden nodes should be manually tuned, for better learning and generalization of ELM. In this paper, we propose a Sparse Semi-Supervised Extreme Learning Machine (S3ELM) via joint sparse regularization for classification, which can automatically prune the model structure via joint sparse regularization technology, to achieve more accurate, efficient and robust classification, when only a small number of labeled training samples are available. Different with most of greedy-algorithms based model selection approaches, by using ℓ2,1-norm, S3ELM casts a joint sparse constraints on the training model of ELM and formulate a convex programming. Moreover, with a Laplacian, S3ELM can make full use of the information from both the labeled and unlabeled samples. Some experiments are taken on several benchmark datasets, and the results show that S3ELM is computationally attractive and outperforms its counterparts.

KW - Extreme learning machine

KW - Joint sparse regularization

KW - Laplacian

KW - Sparse semi-supervised learning

KW - ℓ -Norm

UR - http://www.scopus.com/inward/record.url?scp=84915822678&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84915822678&partnerID=8YFLogxK

U2 - 10.1016/j.knosys.2014.09.014

DO - 10.1016/j.knosys.2014.09.014

M3 - Article

AN - SCOPUS:84915822678

VL - 73

SP - 149

EP - 160

JO - Knowledge-Based Systems

JF - Knowledge-Based Systems

SN - 0950-7051

ER -