Large Margin Distribution Machine Recursive Feature Elimination

Research output: Chapter in Book/Report/Conference proceedingConference contribution

13 Downloads (Pure)

Abstract

—In order to eliminate irrelevant features for
classification, we propose a novel feature selection algorithm
called Large Margin Distribution Machine Recursive Feature
Elimination (LDM-RFE). LDM-RFE uses the latest support
vector based classification algorithm Large Margin Distribution
Machine (LDM) to evaluate all the features of samples, and then
generates a ranked feature list during the procedure of Recursive
Feature Elimination (RFE). In the experiment section, we report
promising results obtained by LDM-RFE in comparison with
several common feature selection algorithms on five UCI
benchmark datasets.
Original languageEnglish
Title of host publicationThe 2017 4th International Conference on Systems and Informatics (ICSAI 2017)
PublisherIEEE Press
Pages1427-1432
Number of pages6
ISBN (Print)978-1-5386-1106-7
Publication statusPublished - 2017
EventThe 2017 4th International Conference on Systems and Informatics - Hangzhou Zhejiang, China
Duration: 11 Nov 201713 Nov 2017

Conference

ConferenceThe 2017 4th International Conference on Systems and Informatics
Abbreviated titleICSAI 2017
CountryChina
CityHangzhou Zhejiang
Period11/11/1713/11/17

Fingerprint

Feature extraction
Experiments

Keywords

  • large margin distribution machine
  • recursive feature elimination
  • classification

Cite this

Ou, G., Wang, Y., Pang, W., & Coghill, G. M. (2017). Large Margin Distribution Machine Recursive Feature Elimination. In The 2017 4th International Conference on Systems and Informatics (ICSAI 2017) (pp. 1427-1432). IEEE Press.

Large Margin Distribution Machine Recursive Feature Elimination. / Ou, Ge; Wang, Yan; Pang, Wei; Coghill, George MacLeod.

The 2017 4th International Conference on Systems and Informatics (ICSAI 2017) . IEEE Press, 2017. p. 1427-1432.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ou, G, Wang, Y, Pang, W & Coghill, GM 2017, Large Margin Distribution Machine Recursive Feature Elimination. in The 2017 4th International Conference on Systems and Informatics (ICSAI 2017) . IEEE Press, pp. 1427-1432, The 2017 4th International Conference on Systems and Informatics , Hangzhou Zhejiang, China, 11/11/17.
Ou G, Wang Y, Pang W, Coghill GM. Large Margin Distribution Machine Recursive Feature Elimination. In The 2017 4th International Conference on Systems and Informatics (ICSAI 2017) . IEEE Press. 2017. p. 1427-1432
Ou, Ge ; Wang, Yan ; Pang, Wei ; Coghill, George MacLeod. / Large Margin Distribution Machine Recursive Feature Elimination. The 2017 4th International Conference on Systems and Informatics (ICSAI 2017) . IEEE Press, 2017. pp. 1427-1432
@inproceedings{66424034a0544396ae462fdd980ed58a,
title = "Large Margin Distribution Machine Recursive Feature Elimination",
abstract = "—In order to eliminate irrelevant features forclassification, we propose a novel feature selection algorithmcalled Large Margin Distribution Machine Recursive FeatureElimination (LDM-RFE). LDM-RFE uses the latest supportvector based classification algorithm Large Margin DistributionMachine (LDM) to evaluate all the features of samples, and thengenerates a ranked feature list during the procedure of RecursiveFeature Elimination (RFE). In the experiment section, we reportpromising results obtained by LDM-RFE in comparison withseveral common feature selection algorithms on five UCIbenchmark datasets.",
keywords = "large margin distribution machine, recursive feature elimination, classification",
author = "Ge Ou and Yan Wang and Wei Pang and Coghill, {George MacLeod}",
note = "We gratefully thank Dr Teng Zhang and Prof Zhi-Hua Zhou for providing the source code of “LDM” source code and their kind technical assistance. This work is supported by the National Natural Science Foundation of China (Nos. 61472159, 61572227) and Development Project of Jilin Province of China (Nos. 20160204022GX, 2017C033). This work is also partially supported by the 2015 Scottish Crucible Award funded by the Royal Society of Edinburgh and the 2016 PECE bursary provided by the Scottish Informatics & Computer Science Alliance (SICSA).",
year = "2017",
language = "English",
isbn = "978-1-5386-1106-7",
pages = "1427--1432",
booktitle = "The 2017 4th International Conference on Systems and Informatics (ICSAI 2017)",
publisher = "IEEE Press",

}

TY - GEN

T1 - Large Margin Distribution Machine Recursive Feature Elimination

AU - Ou, Ge

AU - Wang, Yan

AU - Pang, Wei

AU - Coghill, George MacLeod

N1 - We gratefully thank Dr Teng Zhang and Prof Zhi-Hua Zhou for providing the source code of “LDM” source code and their kind technical assistance. This work is supported by the National Natural Science Foundation of China (Nos. 61472159, 61572227) and Development Project of Jilin Province of China (Nos. 20160204022GX, 2017C033). This work is also partially supported by the 2015 Scottish Crucible Award funded by the Royal Society of Edinburgh and the 2016 PECE bursary provided by the Scottish Informatics & Computer Science Alliance (SICSA).

PY - 2017

Y1 - 2017

N2 - —In order to eliminate irrelevant features forclassification, we propose a novel feature selection algorithmcalled Large Margin Distribution Machine Recursive FeatureElimination (LDM-RFE). LDM-RFE uses the latest supportvector based classification algorithm Large Margin DistributionMachine (LDM) to evaluate all the features of samples, and thengenerates a ranked feature list during the procedure of RecursiveFeature Elimination (RFE). In the experiment section, we reportpromising results obtained by LDM-RFE in comparison withseveral common feature selection algorithms on five UCIbenchmark datasets.

AB - —In order to eliminate irrelevant features forclassification, we propose a novel feature selection algorithmcalled Large Margin Distribution Machine Recursive FeatureElimination (LDM-RFE). LDM-RFE uses the latest supportvector based classification algorithm Large Margin DistributionMachine (LDM) to evaluate all the features of samples, and thengenerates a ranked feature list during the procedure of RecursiveFeature Elimination (RFE). In the experiment section, we reportpromising results obtained by LDM-RFE in comparison withseveral common feature selection algorithms on five UCIbenchmark datasets.

KW - large margin distribution machine

KW - recursive feature elimination

KW - classification

M3 - Conference contribution

SN - 978-1-5386-1106-7

SP - 1427

EP - 1432

BT - The 2017 4th International Conference on Systems and Informatics (ICSAI 2017)

PB - IEEE Press

ER -