Combining feature ranking with PCA

an application to gait analysis

Ming-Jing Yang, Hui-Ru Zheng, Hai-Ying Wang, Sally McClean, Nigel Harris

Research output: Chapter in Book/Report/Conference proceedingChapter

6 Citations (Scopus)

Abstract

Feature reduction is an effective way to improve the classification performance when machine learning methods are used in gait analysis. In this paper, we proposed a novel hybrid feature reduction method (MSNRPCA) based on the combination of feature ranking with principle component analysis (PCA). Three feature reduction methods, namely, feature ranking based the value of signal to noise ratio (MSNR), PCA and the proposed hybrid approach (MSNRPCA), were examined in two gait analysis problems. One gait analysis problem is to differentiate the patients with Neurodegenerative disease from the controls based on the gait data collected by footswitches. The other problem is to discriminate the patients with complex regional pain syndrome (CRPS) from controls based on the gait data collected by an accelerometer. Results showed that the proposed MSNRPCA achieved best classification performance in two gait datasets. In footswitch data, the highest accuracy (81.78%) was obtained using a feature subset with 4 features generated from original 10 features by MSNRPCA. In the accelerometer dataset, classification with three features generated from 17 features by MSNRPCA achieved the best performance with an accuracy of 100%.
Original languageEnglish
Title of host publication2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010
Place of PublicationPiscataway, U. S. A.
PublisherIEEE
Pages494-499
Number of pages6
ISBN (Electronic)978-1-4244-6527-9
ISBN (Print)9781424465262
DOIs
Publication statusPublished - Jul 2010
Event2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010, July 11, 2010 - July 14, 2010 - Qingdao, China
Duration: 1 Jul 2010 → …

Conference

Conference2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010, July 11, 2010 - July 14, 2010
CountryChina
CityQingdao
Period1/07/10 → …

Fingerprint

Gait analysis
Accelerometers
Neurodegenerative diseases
Learning systems
Signal to noise ratio

Cite this

Yang, M-J., Zheng, H-R., Wang, H-Y., McClean, S., & Harris, N. (2010). Combining feature ranking with PCA: an application to gait analysis. In 2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010 (pp. 494-499). [5581013] Piscataway, U. S. A.: IEEE. https://doi.org/10.1109/ICMLC.2010.5581013

Combining feature ranking with PCA : an application to gait analysis. / Yang, Ming-Jing; Zheng, Hui-Ru; Wang, Hai-Ying; McClean, Sally; Harris, Nigel.

2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010. Piscataway, U. S. A. : IEEE, 2010. p. 494-499 5581013.

Research output: Chapter in Book/Report/Conference proceedingChapter

Yang, M-J, Zheng, H-R, Wang, H-Y, McClean, S & Harris, N 2010, Combining feature ranking with PCA: an application to gait analysis. in 2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010., 5581013, IEEE, Piscataway, U. S. A., pp. 494-499, 2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010, July 11, 2010 - July 14, 2010, Qingdao, China, 1/07/10. https://doi.org/10.1109/ICMLC.2010.5581013
Yang M-J, Zheng H-R, Wang H-Y, McClean S, Harris N. Combining feature ranking with PCA: an application to gait analysis. In 2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010. Piscataway, U. S. A.: IEEE. 2010. p. 494-499. 5581013 https://doi.org/10.1109/ICMLC.2010.5581013
Yang, Ming-Jing ; Zheng, Hui-Ru ; Wang, Hai-Ying ; McClean, Sally ; Harris, Nigel. / Combining feature ranking with PCA : an application to gait analysis. 2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010. Piscataway, U. S. A. : IEEE, 2010. pp. 494-499
@inbook{5634cb9772e04f9196bcd91e693d3916,
title = "Combining feature ranking with PCA: an application to gait analysis",
abstract = "Feature reduction is an effective way to improve the classification performance when machine learning methods are used in gait analysis. In this paper, we proposed a novel hybrid feature reduction method (MSNRPCA) based on the combination of feature ranking with principle component analysis (PCA). Three feature reduction methods, namely, feature ranking based the value of signal to noise ratio (MSNR), PCA and the proposed hybrid approach (MSNRPCA), were examined in two gait analysis problems. One gait analysis problem is to differentiate the patients with Neurodegenerative disease from the controls based on the gait data collected by footswitches. The other problem is to discriminate the patients with complex regional pain syndrome (CRPS) from controls based on the gait data collected by an accelerometer. Results showed that the proposed MSNRPCA achieved best classification performance in two gait datasets. In footswitch data, the highest accuracy (81.78{\%}) was obtained using a feature subset with 4 features generated from original 10 features by MSNRPCA. In the accelerometer dataset, classification with three features generated from 17 features by MSNRPCA achieved the best performance with an accuracy of 100{\%}.",
author = "Ming-Jing Yang and Hui-Ru Zheng and Hai-Ying Wang and Sally McClean and Nigel Harris",
year = "2010",
month = "7",
doi = "10.1109/ICMLC.2010.5581013",
language = "English",
isbn = "9781424465262",
pages = "494--499",
booktitle = "2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010",
publisher = "IEEE",
address = "USA United States",

}

TY - CHAP

T1 - Combining feature ranking with PCA

T2 - an application to gait analysis

AU - Yang, Ming-Jing

AU - Zheng, Hui-Ru

AU - Wang, Hai-Ying

AU - McClean, Sally

AU - Harris, Nigel

PY - 2010/7

Y1 - 2010/7

N2 - Feature reduction is an effective way to improve the classification performance when machine learning methods are used in gait analysis. In this paper, we proposed a novel hybrid feature reduction method (MSNRPCA) based on the combination of feature ranking with principle component analysis (PCA). Three feature reduction methods, namely, feature ranking based the value of signal to noise ratio (MSNR), PCA and the proposed hybrid approach (MSNRPCA), were examined in two gait analysis problems. One gait analysis problem is to differentiate the patients with Neurodegenerative disease from the controls based on the gait data collected by footswitches. The other problem is to discriminate the patients with complex regional pain syndrome (CRPS) from controls based on the gait data collected by an accelerometer. Results showed that the proposed MSNRPCA achieved best classification performance in two gait datasets. In footswitch data, the highest accuracy (81.78%) was obtained using a feature subset with 4 features generated from original 10 features by MSNRPCA. In the accelerometer dataset, classification with three features generated from 17 features by MSNRPCA achieved the best performance with an accuracy of 100%.

AB - Feature reduction is an effective way to improve the classification performance when machine learning methods are used in gait analysis. In this paper, we proposed a novel hybrid feature reduction method (MSNRPCA) based on the combination of feature ranking with principle component analysis (PCA). Three feature reduction methods, namely, feature ranking based the value of signal to noise ratio (MSNR), PCA and the proposed hybrid approach (MSNRPCA), were examined in two gait analysis problems. One gait analysis problem is to differentiate the patients with Neurodegenerative disease from the controls based on the gait data collected by footswitches. The other problem is to discriminate the patients with complex regional pain syndrome (CRPS) from controls based on the gait data collected by an accelerometer. Results showed that the proposed MSNRPCA achieved best classification performance in two gait datasets. In footswitch data, the highest accuracy (81.78%) was obtained using a feature subset with 4 features generated from original 10 features by MSNRPCA. In the accelerometer dataset, classification with three features generated from 17 features by MSNRPCA achieved the best performance with an accuracy of 100%.

UR - http://dx.doi.org/10.1109/ICMLC.2010.5581013

U2 - 10.1109/ICMLC.2010.5581013

DO - 10.1109/ICMLC.2010.5581013

M3 - Chapter

SN - 9781424465262

SP - 494

EP - 499

BT - 2010 International Conference on Machine Learning and Cybernetics, ICMLC 2010

PB - IEEE

CY - Piscataway, U. S. A.

ER -