Online Semi-Supervised Learning With Multiple Regularization Terms
Online semi-supervised learning (OS<sup>2</sup>L) has received much attention recently because of its well practical usefulness. Most of the existing studies of OS<sup>2</sup>L are related to manifold regularization. In this paper, we introduce a novel OS<sup>2</sup&...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2019-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/8633907/ |
id |
doaj-987ac0075b094213a93ba13db4fce34f |
---|---|
record_format |
Article |
spelling |
doaj-987ac0075b094213a93ba13db4fce34f2021-03-29T23:38:02ZengIEEEIEEE Access2169-35362019-01-017684796849410.1109/ACCESS.2019.28973828633907Online Semi-Supervised Learning With Multiple Regularization TermsChao Chen0Boliang Sun1Xingchen Hu2Yan Li3College of Systems Engineering, National University of Defense Technology, Changsha, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha, ChinaCollege of Systems Engineering, National University of Defense Technology, Changsha, ChinaOnline semi-supervised learning (OS<sup>2</sup>L) has received much attention recently because of its well practical usefulness. Most of the existing studies of OS<sup>2</sup>L are related to manifold regularization. In this paper, we introduce a novel OS<sup>2</sup>L framework with multiple regularization terms based on the notion of ascending the dual function in constrained optimization. Using the Fenchel conjugate, different semi-supervised regularization terms can be integrated into the dual function easily and directly. This approach is derived by updating limited dual coefficient variables on each learning round. To be practical, we also employ buffering strategies and sparse approximation approaches in this paper. The experimental studies show that our methods achieve accuracy comparable to offline algorithms while consuming less time and memory. Especially, our OS<sup>2</sup>L algorithms can handle the settings where the target hyperplane of classification continually drifts with the sequence of arriving instances. This paper paves a way to design and analyze OS<sup>2</sup>L algorithms with multiple regularization terms.https://ieeexplore.ieee.org/document/8633907/Online semi-supervised learning (OS²L)SVMmanifold regularizationco-regularizationFenchel conjugate |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Chao Chen Boliang Sun Xingchen Hu Yan Li |
spellingShingle |
Chao Chen Boliang Sun Xingchen Hu Yan Li Online Semi-Supervised Learning With Multiple Regularization Terms IEEE Access Online semi-supervised learning (OS²L) SVM manifold regularization co-regularization Fenchel conjugate |
author_facet |
Chao Chen Boliang Sun Xingchen Hu Yan Li |
author_sort |
Chao Chen |
title |
Online Semi-Supervised Learning With Multiple Regularization Terms |
title_short |
Online Semi-Supervised Learning With Multiple Regularization Terms |
title_full |
Online Semi-Supervised Learning With Multiple Regularization Terms |
title_fullStr |
Online Semi-Supervised Learning With Multiple Regularization Terms |
title_full_unstemmed |
Online Semi-Supervised Learning With Multiple Regularization Terms |
title_sort |
online semi-supervised learning with multiple regularization terms |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2019-01-01 |
description |
Online semi-supervised learning (OS<sup>2</sup>L) has received much attention recently because of its well practical usefulness. Most of the existing studies of OS<sup>2</sup>L are related to manifold regularization. In this paper, we introduce a novel OS<sup>2</sup>L framework with multiple regularization terms based on the notion of ascending the dual function in constrained optimization. Using the Fenchel conjugate, different semi-supervised regularization terms can be integrated into the dual function easily and directly. This approach is derived by updating limited dual coefficient variables on each learning round. To be practical, we also employ buffering strategies and sparse approximation approaches in this paper. The experimental studies show that our methods achieve accuracy comparable to offline algorithms while consuming less time and memory. Especially, our OS<sup>2</sup>L algorithms can handle the settings where the target hyperplane of classification continually drifts with the sequence of arriving instances. This paper paves a way to design and analyze OS<sup>2</sup>L algorithms with multiple regularization terms. |
topic |
Online semi-supervised learning (OS²L) SVM manifold regularization co-regularization Fenchel conjugate |
url |
https://ieeexplore.ieee.org/document/8633907/ |
work_keys_str_mv |
AT chaochen onlinesemisupervisedlearningwithmultipleregularizationterms AT boliangsun onlinesemisupervisedlearningwithmultipleregularizationterms AT xingchenhu onlinesemisupervisedlearningwithmultipleregularizationterms AT yanli onlinesemisupervisedlearningwithmultipleregularizationterms |
_version_ |
1724189222291636224 |