Bibliothèque centrale Université Mustapha Stambouli de Mascara

Lien vers la notice

Titre :	EEG signal classification using ICA and machine learning
Auteurs :	Karima Diaf, Auteur ; Meriem Aribi, Auteur ; Brahim Cherouati, Directeur de thèse
Type de document :	texte manuscrit
Editeur :	Université mustapha stambouli de Mascara:Faculté des sciences exactes, 2023
ISBN/ISSN/EAN :	SE02335T
Format :	72P. / couv. ill. / 29cm.
Accompagnement :	disque optique numérique (CD-ROM)
Langues:	Anglais
Résumé :	In this work, we have tackled a problem of binary classification in the field of neurophysiologic diseases applied to epilepsy; works has been carried out in this field of research using the EEG technology. To carry out this work, we used data supervised by a neurophysiologist, these data are recorded using an EEG, and we carried out a set of tests on these data, interesting results are obtained compared with some of conventional methods. In order to automate the classification procedure, two classification methods were used, namely SVM and MLP. Our aim was to highlight these classification methods through validation and performance evaluation tests. Good results were observed using frequency features studied with the MLP classification method. This study also showed that the SVM classification method is better adapted to the data than the MLP method. Finally, we believe that our approach can be used for medical purposes for offline data, with the future aim of automating this approach for online data, and exploring other classification methods