Non-linear classifiers applied to EEG analysis for epilepsy seizure detection

•BoW and non-linear SVM, improve accuracy rates found in literature.•BoW-inspired method is very convenient for noisy signals.•The use of non-linear kernels for BoW reveals unnecessary.•The BoW performance is due to the more linear and discriminative space it creates. This work presents a novel appr...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	Expert systems with applications 2017-11, Vol.86, p.99-112
Hauptverfasser:	Martinez-del-Rincon, Jesus, Santofimia, Maria J., del Toro, Xavier, Barba, Jesus, Romero, Francisca, Navas, Patricia, Lopez, Juan C.
Format:	Artikel
Sprache:	eng
Schlagworte:	Algorithms Bag of words Classification Classification algorithms Classifiers Convulsions & seizures Datasets Electroencephalography Epilepsy Feature extraction Non-linear classifiers Robustness (mathematics) SVM Wavelet
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!