Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal

Saugat Bhattacharyya; Amit Konar; D. N. Tibarewala; Anwesha Khasnobish; R. Janarthanan

doi:10.1109/CIEC.2014.6959183

Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal

Saugat Bhattacharyya, Amit Konar, D. N. Tibarewala, Anwesha Khasnobish, R. Janarthanan

School of Computing, Eng & Intel. Sys

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

20 Citations (Scopus)

163 Downloads (Pure)

Abstract

Recent advances in the field of Brain-computer Interfacing (BCI) has opened wide potentials in neuro-rehabilitative applications. Electeroencephalography (EEG) is the most frequently used brain measure in BCI research. Mental states are distinguished from classifiers which uses features extracted from the raw EEG as inputs. Ensemble classifiers combine a number of classifiers or learners to improve the classification results. It is more suited for multi-class classification of time-varying EEG signal. In this paper, we have used AdaBoost, LPBoost, RUSBoost, Bagging and Random Subspaces for classification of 3-class motor imagery EEG data. For this purpose, we have employed adaptive autoregressive coefficients as features and feed forward neural network (FFNN) as the base learner of the ensemble methods. The results show that the classification accuracies of the ensemble classifiers except RUSBoost performs better than a single FFNN classifier.

Original language	English
Title of host publication	International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	712-716
Number of pages	5
ISBN (Electronic)	9781479920440
DOIs	https://doi.org/10.1109/CIEC.2014.6959183
Publication status	Published (in print/issue) - 14 Nov 2014
Event	2014 International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014 - Kolkata, India Duration: 31 Jan 2014 → 2 Feb 2014

Publication series

Name	International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014

Conference

Conference	2014 International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014
Country/Territory	India
City	Kolkata
Period	31/01/14 → 2/02/14

Keywords

Adaptive Autoregressive Parameter
Electroencephalography
Ensemble methods
Feed Forward Neural Network
Motor imagery
Multi-class classification

Access to Document

10.1109/CIEC.2014.6959183

paperAuthor Accepted Manuscript, 308 KB

Cite this

Bhattacharyya, S., Konar, A., Tibarewala, D. N., Khasnobish, A., & Janarthanan, R. (2014). Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal. In International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014 (pp. 712-716). Article 6959183 (International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CIEC.2014.6959183

Bhattacharyya, Saugat ; Konar, Amit ; Tibarewala, D. N. et al. / Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal. International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. pp. 712-716 (International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014).

@inproceedings{b590f708bb0c4e6bb39b3ce92ea0a0b0,

title = "Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal",

abstract = "Recent advances in the field of Brain-computer Interfacing (BCI) has opened wide potentials in neuro-rehabilitative applications. Electeroencephalography (EEG) is the most frequently used brain measure in BCI research. Mental states are distinguished from classifiers which uses features extracted from the raw EEG as inputs. Ensemble classifiers combine a number of classifiers or learners to improve the classification results. It is more suited for multi-class classification of time-varying EEG signal. In this paper, we have used AdaBoost, LPBoost, RUSBoost, Bagging and Random Subspaces for classification of 3-class motor imagery EEG data. For this purpose, we have employed adaptive autoregressive coefficients as features and feed forward neural network (FFNN) as the base learner of the ensemble methods. The results show that the classification accuracies of the ensemble classifiers except RUSBoost performs better than a single FFNN classifier.",

keywords = "Adaptive Autoregressive Parameter, Electroencephalography, Ensemble methods, Feed Forward Neural Network, Motor imagery, Multi-class classification",

author = "Saugat Bhattacharyya and Amit Konar and Tibarewala, {D. N.} and Anwesha Khasnobish and R. Janarthanan",

year = "2014",

month = nov,

day = "14",

doi = "10.1109/CIEC.2014.6959183",

language = "English",

series = "International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "712--716",

booktitle = "International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014",

address = "United States",

note = "2014 International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014 ; Conference date: 31-01-2014 Through 02-02-2014",

}

Bhattacharyya, S, Konar, A, Tibarewala, DN, Khasnobish, A & Janarthanan, R 2014, Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal. in International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014., 6959183, International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014, Institute of Electrical and Electronics Engineers Inc., pp. 712-716, 2014 International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014, Kolkata, India, 31/01/14. https://doi.org/10.1109/CIEC.2014.6959183

Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal. / Bhattacharyya, Saugat; Konar, Amit; Tibarewala, D. N. et al.
International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014. Institute of Electrical and Electronics Engineers Inc., 2014. p. 712-716 6959183 (International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal

AU - Bhattacharyya, Saugat

AU - Konar, Amit

AU - Tibarewala, D. N.

AU - Khasnobish, Anwesha

AU - Janarthanan, R.

PY - 2014/11/14

Y1 - 2014/11/14

N2 - Recent advances in the field of Brain-computer Interfacing (BCI) has opened wide potentials in neuro-rehabilitative applications. Electeroencephalography (EEG) is the most frequently used brain measure in BCI research. Mental states are distinguished from classifiers which uses features extracted from the raw EEG as inputs. Ensemble classifiers combine a number of classifiers or learners to improve the classification results. It is more suited for multi-class classification of time-varying EEG signal. In this paper, we have used AdaBoost, LPBoost, RUSBoost, Bagging and Random Subspaces for classification of 3-class motor imagery EEG data. For this purpose, we have employed adaptive autoregressive coefficients as features and feed forward neural network (FFNN) as the base learner of the ensemble methods. The results show that the classification accuracies of the ensemble classifiers except RUSBoost performs better than a single FFNN classifier.

AB - Recent advances in the field of Brain-computer Interfacing (BCI) has opened wide potentials in neuro-rehabilitative applications. Electeroencephalography (EEG) is the most frequently used brain measure in BCI research. Mental states are distinguished from classifiers which uses features extracted from the raw EEG as inputs. Ensemble classifiers combine a number of classifiers or learners to improve the classification results. It is more suited for multi-class classification of time-varying EEG signal. In this paper, we have used AdaBoost, LPBoost, RUSBoost, Bagging and Random Subspaces for classification of 3-class motor imagery EEG data. For this purpose, we have employed adaptive autoregressive coefficients as features and feed forward neural network (FFNN) as the base learner of the ensemble methods. The results show that the classification accuracies of the ensemble classifiers except RUSBoost performs better than a single FFNN classifier.

KW - Adaptive Autoregressive Parameter

KW - Electroencephalography

KW - Ensemble methods

KW - Feed Forward Neural Network

KW - Motor imagery

KW - Multi-class classification

UR - http://www.scopus.com/inward/record.url?scp=84914689386&partnerID=8YFLogxK

U2 - 10.1109/CIEC.2014.6959183

DO - 10.1109/CIEC.2014.6959183

M3 - Conference contribution

AN - SCOPUS:84914689386

T3 - International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014

SP - 712

EP - 716

BT - International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2014 International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014

Y2 - 31 January 2014 through 2 February 2014

ER -

Bhattacharyya S, Konar A, Tibarewala DN, Khasnobish A, Janarthanan R. Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal. In International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014. Institute of Electrical and Electronics Engineers Inc. 2014. p. 712-716. 6959183. (International Conference on Control, Instrumentation, Energy and Communication, CIEC 2014). doi: 10.1109/CIEC.2014.6959183

Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Motion control of artificial limb(s) through a human-computer interface

Cite this

Performance analysis of ensemble methods for multi-class classification of motor imagery EEG signal

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Prizes

Motion control of artificial limb(s) through a human-computer interface

Cite this