Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface

Giovanni Saggio, Pietro Cavallo, Giovanni Costantini, Gianluca Susi, Lucia Rita Quitadamo, Maria Grazia Marciani, Luigi Bianchi

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

Abstract

In this study a comparison among three different machine learning techniques for the classification of mental tasks for a Brain-Computer Interface system is presented: MLP neural network, Fuzzy C-Means Analysis and Support Vector Machine (SVM). In BCI literature, finding the best classifier is a very hard problem to solve, and it is still an open question. We considered only ten electrodes for our analysis, in order to lower the computational workload. Different parameters were analyzed for the evaluation of the performances of the classifiers: accuracy, training time and size of the training dataset. Results demonstrated how the accuracies of the three classifiers are nearly the same but the error margin of SVM on this reduced dataset is larger compared to the other two classifiers. Furthermore neural network needs a reduced number of trials for training purposes, reducing the recording session up to 8 times with respect to SVM and Fuzzy analysis. This suggests how, in the presented case, MLP neural network can be preferable for the classification of mental tasks in Brain Computer Interface systems.

Original languageEnglish
Title of host publicationBIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings
Pages174-179
Number of pages6
Publication statusPublished - 2010
Event3rd International Conference on Bio-inspired Systems and Signal Processing, BIOSIGNALS 2010 - Valencia, Spain
Duration: Jan 20 2010Jan 23 2010

Other

Other3rd International Conference on Bio-inspired Systems and Signal Processing, BIOSIGNALS 2010
CountrySpain
CityValencia
Period1/20/101/23/10

Fingerprint

Brain computer interface
Classifiers
Support vector machines
Neural networks
Fuzzy neural networks
Learning systems
Electrodes

Keywords

  • BCI
  • Fuzzy logic
  • Neural networks
  • SVM

ASJC Scopus subject areas

  • Signal Processing

Cite this

Saggio, G., Cavallo, P., Costantini, G., Susi, G., Quitadamo, L. R., Marciani, M. G., & Bianchi, L. (2010). Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface. In BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings (pp. 174-179)

Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface. / Saggio, Giovanni; Cavallo, Pietro; Costantini, Giovanni; Susi, Gianluca; Quitadamo, Lucia Rita; Marciani, Maria Grazia; Bianchi, Luigi.

BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings. 2010. p. 174-179.

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

Saggio, G, Cavallo, P, Costantini, G, Susi, G, Quitadamo, LR, Marciani, MG & Bianchi, L 2010, Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface. in BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings. pp. 174-179, 3rd International Conference on Bio-inspired Systems and Signal Processing, BIOSIGNALS 2010, Valencia, Spain, 1/20/10.
Saggio G, Cavallo P, Costantini G, Susi G, Quitadamo LR, Marciani MG et al. Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface. In BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings. 2010. p. 174-179
Saggio, Giovanni ; Cavallo, Pietro ; Costantini, Giovanni ; Susi, Gianluca ; Quitadamo, Lucia Rita ; Marciani, Maria Grazia ; Bianchi, Luigi. / Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface. BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings. 2010. pp. 174-179
@inproceedings{a4124342b08a42569ef155442be8cb3d,
title = "Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface",
abstract = "In this study a comparison among three different machine learning techniques for the classification of mental tasks for a Brain-Computer Interface system is presented: MLP neural network, Fuzzy C-Means Analysis and Support Vector Machine (SVM). In BCI literature, finding the best classifier is a very hard problem to solve, and it is still an open question. We considered only ten electrodes for our analysis, in order to lower the computational workload. Different parameters were analyzed for the evaluation of the performances of the classifiers: accuracy, training time and size of the training dataset. Results demonstrated how the accuracies of the three classifiers are nearly the same but the error margin of SVM on this reduced dataset is larger compared to the other two classifiers. Furthermore neural network needs a reduced number of trials for training purposes, reducing the recording session up to 8 times with respect to SVM and Fuzzy analysis. This suggests how, in the presented case, MLP neural network can be preferable for the classification of mental tasks in Brain Computer Interface systems.",
keywords = "BCI, Fuzzy logic, Neural networks, SVM",
author = "Giovanni Saggio and Pietro Cavallo and Giovanni Costantini and Gianluca Susi and Quitadamo, {Lucia Rita} and Marciani, {Maria Grazia} and Luigi Bianchi",
year = "2010",
language = "English",
isbn = "9789896740184",
pages = "174--179",
booktitle = "BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings",

}

TY - GEN

T1 - Comparison of different classifiers on a reduced set of features for mental tasks-based brain computer interface

AU - Saggio, Giovanni

AU - Cavallo, Pietro

AU - Costantini, Giovanni

AU - Susi, Gianluca

AU - Quitadamo, Lucia Rita

AU - Marciani, Maria Grazia

AU - Bianchi, Luigi

PY - 2010

Y1 - 2010

N2 - In this study a comparison among three different machine learning techniques for the classification of mental tasks for a Brain-Computer Interface system is presented: MLP neural network, Fuzzy C-Means Analysis and Support Vector Machine (SVM). In BCI literature, finding the best classifier is a very hard problem to solve, and it is still an open question. We considered only ten electrodes for our analysis, in order to lower the computational workload. Different parameters were analyzed for the evaluation of the performances of the classifiers: accuracy, training time and size of the training dataset. Results demonstrated how the accuracies of the three classifiers are nearly the same but the error margin of SVM on this reduced dataset is larger compared to the other two classifiers. Furthermore neural network needs a reduced number of trials for training purposes, reducing the recording session up to 8 times with respect to SVM and Fuzzy analysis. This suggests how, in the presented case, MLP neural network can be preferable for the classification of mental tasks in Brain Computer Interface systems.

AB - In this study a comparison among three different machine learning techniques for the classification of mental tasks for a Brain-Computer Interface system is presented: MLP neural network, Fuzzy C-Means Analysis and Support Vector Machine (SVM). In BCI literature, finding the best classifier is a very hard problem to solve, and it is still an open question. We considered only ten electrodes for our analysis, in order to lower the computational workload. Different parameters were analyzed for the evaluation of the performances of the classifiers: accuracy, training time and size of the training dataset. Results demonstrated how the accuracies of the three classifiers are nearly the same but the error margin of SVM on this reduced dataset is larger compared to the other two classifiers. Furthermore neural network needs a reduced number of trials for training purposes, reducing the recording session up to 8 times with respect to SVM and Fuzzy analysis. This suggests how, in the presented case, MLP neural network can be preferable for the classification of mental tasks in Brain Computer Interface systems.

KW - BCI

KW - Fuzzy logic

KW - Neural networks

KW - SVM

UR - http://www.scopus.com/inward/record.url?scp=77956429806&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77956429806&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9789896740184

SP - 174

EP - 179

BT - BIOSIGNALS 2010 - Proceedings of the 3rd International Conference on Bio-inpsired Systems and Signal Processing, Proceedings

ER -