A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints

Previous studies of Non-Parametric Kernel Learning (NPKL) usually formulate the learning task as a Semi-Definite Programming (SDP) problem that is often solved by some general purpose SDP solvers. However, for N data examples, the time complexity of NPKL using a standard interior-point SDP solver co...

Full description

Saved in:
Bibliographic Details
Main Authors: ZHUANG, Jinfeng, TSANG, Ivor W., HOI, Steven C. H.
Format: text
Language:English
Published: Institutional Knowledge at Singapore Management University 2011
Subjects:
Online Access:https://ink.library.smu.edu.sg/sis_research/2289
https://ink.library.smu.edu.sg/context/sis_research/article/3289/viewcontent/A_Family_of_Simple_Non_Parametric_Kernel_Learning_Algorithms_from_Pairwise_Constraints.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Singapore Management University
Language: English
id sg-smu-ink.sis_research-3289
record_format dspace
spelling sg-smu-ink.sis_research-32892018-12-06T03:43:52Z A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints ZHUANG, Jinfeng TSANG, Ivor W. HOI, Steven C. H. Previous studies of Non-Parametric Kernel Learning (NPKL) usually formulate the learning task as a Semi-Definite Programming (SDP) problem that is often solved by some general purpose SDP solvers. However, for N data examples, the time complexity of NPKL using a standard interior-point SDP solver could be as high as O(N6.5), which prohibits NPKL methods applicable to real applications, even for data sets of moderate size. In this paper, we present a family of efficient NPKL algorithms, termed "SimpleNPKL", which can learn non-parametric kernels from a large set of pairwise constraints efficiently. In particular, we propose two efficient SimpleNPKL algorithms. One is SimpleNPKL algorithm with linear loss, which enjoys a closed-form solution that can be efficiently computed by the Lanczos sparse eigen decomposition technique. Another one is SimpleNPKL algorithm with other loss functions (including square hinge loss, hinge loss, square loss) that can be re-formulated as a saddle-point optimization problem, which can be further resolved by a fast iterative algorithm. In contrast to the previous NPKL approaches, our empirical results show that the proposed new technique, maintaining the same accuracy, is significantly more efficient and scalable. Finally, we also demonstrate that the proposed new technique is also applicable to speed up many kernel learning tasks, including colored maximum variance unfolding, minimum volume embedding, and structure preserving embedding. 2011-04-01T07:00:00Z text application/pdf https://ink.library.smu.edu.sg/sis_research/2289 https://ink.library.smu.edu.sg/context/sis_research/article/3289/viewcontent/A_Family_of_Simple_Non_Parametric_Kernel_Learning_Algorithms_from_Pairwise_Constraints.pdf http://creativecommons.org/licenses/by-nc-nd/4.0/ Research Collection School Of Computing and Information Systems eng Institutional Knowledge at Singapore Management University non-parametric kernel learning pairwise constraints semi-definite programming semi-supervised learning side information Computer Sciences Databases and Information Systems Theory and Algorithms
institution Singapore Management University
building SMU Libraries
continent Asia
country Singapore
Singapore
content_provider SMU Libraries
collection InK@SMU
language English
topic non-parametric kernel learning
pairwise constraints
semi-definite programming
semi-supervised learning
side information
Computer Sciences
Databases and Information Systems
Theory and Algorithms
spellingShingle non-parametric kernel learning
pairwise constraints
semi-definite programming
semi-supervised learning
side information
Computer Sciences
Databases and Information Systems
Theory and Algorithms
ZHUANG, Jinfeng
TSANG, Ivor W.
HOI, Steven C. H.
A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints
description Previous studies of Non-Parametric Kernel Learning (NPKL) usually formulate the learning task as a Semi-Definite Programming (SDP) problem that is often solved by some general purpose SDP solvers. However, for N data examples, the time complexity of NPKL using a standard interior-point SDP solver could be as high as O(N6.5), which prohibits NPKL methods applicable to real applications, even for data sets of moderate size. In this paper, we present a family of efficient NPKL algorithms, termed "SimpleNPKL", which can learn non-parametric kernels from a large set of pairwise constraints efficiently. In particular, we propose two efficient SimpleNPKL algorithms. One is SimpleNPKL algorithm with linear loss, which enjoys a closed-form solution that can be efficiently computed by the Lanczos sparse eigen decomposition technique. Another one is SimpleNPKL algorithm with other loss functions (including square hinge loss, hinge loss, square loss) that can be re-formulated as a saddle-point optimization problem, which can be further resolved by a fast iterative algorithm. In contrast to the previous NPKL approaches, our empirical results show that the proposed new technique, maintaining the same accuracy, is significantly more efficient and scalable. Finally, we also demonstrate that the proposed new technique is also applicable to speed up many kernel learning tasks, including colored maximum variance unfolding, minimum volume embedding, and structure preserving embedding.
format text
author ZHUANG, Jinfeng
TSANG, Ivor W.
HOI, Steven C. H.
author_facet ZHUANG, Jinfeng
TSANG, Ivor W.
HOI, Steven C. H.
author_sort ZHUANG, Jinfeng
title A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints
title_short A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints
title_full A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints
title_fullStr A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints
title_full_unstemmed A Family of Simple Non-Parametric Kernel Learning Algorithms from Pairwise Constraints
title_sort family of simple non-parametric kernel learning algorithms from pairwise constraints
publisher Institutional Knowledge at Singapore Management University
publishDate 2011
url https://ink.library.smu.edu.sg/sis_research/2289
https://ink.library.smu.edu.sg/context/sis_research/article/3289/viewcontent/A_Family_of_Simple_Non_Parametric_Kernel_Learning_Algorithms_from_Pairwise_Constraints.pdf
_version_ 1770572073965453312