Active learning based on support vector machines

Ran WANG; Sam KWONG; Qiang HE

doi:10.1109/ICSMC.2010.5642440

Active learning based on support vector machines

Ran WANG, Sam KWONG, Qiang HE

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

1 Citation (Scopus)

Abstract

Active learning is mainly to select a part of unlabelled samples from a big dataset. The selected samples are then submitted to domain experts to label and added to the training set. Suppose that the price of labeling samples is far more than the computational cost of training algorithms, we propose a scheme of active learning based on support vector machines, which follows the traditionally inductive learning model of general-specific. In terms of the number of selected samples, the training cost, and the generalization ability, a comparison with some existing active learning algorithms is conducted. The advantages and disadvantages are demonstrated experimentally. ©2010 IEEE.

Original language	English
Title of host publication	2010 IEEE International Conference on Systems, Man and Cybernetics
Publisher	IEEE
Pages	1312-1316
Number of pages	5
ISBN (Electronic)	9781424465880
ISBN (Print)	9781424465866
DOIs	https://doi.org/10.1109/ICSMC.2010.5642440
Publication status	Published - Oct 2010
Externally published	Yes
Event	2010 IEEE International Conference on Systems, Man and Cybernetics, SMC 2010 - Istanbul, Turkey Duration: 10 Oct 2010 → 13 Oct 2010

Conference

Conference	2010 IEEE International Conference on Systems, Man and Cybernetics, SMC 2010
Country/Territory	Turkey
City	Istanbul
Period	10/10/10 → 13/10/10

Keywords

Active learning
Order in hypothesis space
Sample selection
SVM

Access to Document

10.1109/ICSMC.2010.5642440

Cite this

@inproceedings{723ea7d2ca934ba480f231b7fed6bdc1,

title = "Active learning based on support vector machines",

abstract = "Active learning is mainly to select a part of unlabelled samples from a big dataset. The selected samples are then submitted to domain experts to label and added to the training set. Suppose that the price of labeling samples is far more than the computational cost of training algorithms, we propose a scheme of active learning based on support vector machines, which follows the traditionally inductive learning model of general-specific. In terms of the number of selected samples, the training cost, and the generalization ability, a comparison with some existing active learning algorithms is conducted. The advantages and disadvantages are demonstrated experimentally. {\textcopyright}2010 IEEE.",

keywords = "Active learning, Order in hypothesis space, Sample selection, SVM",

author = "Ran WANG and Sam KWONG and Qiang HE",

year = "2010",

month = oct,

doi = "10.1109/ICSMC.2010.5642440",

language = "English",

isbn = "9781424465866",

pages = "1312--1316",

booktitle = "2010 IEEE International Conference on Systems, Man and Cybernetics",

publisher = "IEEE",

note = "2010 IEEE International Conference on Systems, Man and Cybernetics, SMC 2010 ; Conference date: 10-10-2010 Through 13-10-2010",

}

TY - GEN

T1 - Active learning based on support vector machines

AU - WANG, Ran

AU - KWONG, Sam

AU - HE, Qiang

PY - 2010/10

Y1 - 2010/10

N2 - Active learning is mainly to select a part of unlabelled samples from a big dataset. The selected samples are then submitted to domain experts to label and added to the training set. Suppose that the price of labeling samples is far more than the computational cost of training algorithms, we propose a scheme of active learning based on support vector machines, which follows the traditionally inductive learning model of general-specific. In terms of the number of selected samples, the training cost, and the generalization ability, a comparison with some existing active learning algorithms is conducted. The advantages and disadvantages are demonstrated experimentally. ©2010 IEEE.

AB - Active learning is mainly to select a part of unlabelled samples from a big dataset. The selected samples are then submitted to domain experts to label and added to the training set. Suppose that the price of labeling samples is far more than the computational cost of training algorithms, we propose a scheme of active learning based on support vector machines, which follows the traditionally inductive learning model of general-specific. In terms of the number of selected samples, the training cost, and the generalization ability, a comparison with some existing active learning algorithms is conducted. The advantages and disadvantages are demonstrated experimentally. ©2010 IEEE.

KW - Active learning

KW - Order in hypothesis space

KW - Sample selection

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=78751500839&partnerID=8YFLogxK

U2 - 10.1109/ICSMC.2010.5642440

DO - 10.1109/ICSMC.2010.5642440

M3 - Conference paper (refereed)

SN - 9781424465866

SP - 1312

EP - 1316

BT - 2010 IEEE International Conference on Systems, Man and Cybernetics

PB - IEEE

T2 - 2010 IEEE International Conference on Systems, Man and Cybernetics, SMC 2010

Y2 - 10 October 2010 through 13 October 2010

ER -

Active learning based on support vector machines

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this