Please use this identifier to cite or link to this item: http://dspace.dtu.ac.in:8080/jspui/handle/repository/15821
Full metadata record
DC FieldValueLanguage
dc.contributor.authorNAGAR, PREETI-
dc.date.accessioned2017-07-24T17:00:21Z-
dc.date.available2017-07-24T17:00:21Z-
dc.date.issued2017-06-
dc.identifier.urihttp://dspace.dtu.ac.in:8080/jspui/handle/repository/15821-
dc.description.abstractIt has been long known that neurons are electrically excitable cells that process and transmit information via electrical and chemical signalling. Electroencephalography (EEG) is the estimation and recording of these electrical signals utilizing sensors exhibited over the scalp. A Brain-Computer Interface(BCI) is an immediate correspondence pathway between the outside gadget and the human cerebrum. The eld of BCI is a primary purpose for utilizing electroencephalography innovation. In the past, the standard focus has been about making applications in a therapeutic setting, helping paralysed or weakened patients to coordinate with the external world mapping brain signals to human intellect and sensory-motor actions. With the innovative work, BCI progress is at no time later on compelled to simply patients or for treatment; there is a move of focus towards general well-being of people. Stress identi cation using EEG signals is one of the critical areas of research in this direction. A high amount of stress is experienced by people of all ages nowadays and is a ecting their physical and mental health adversely. The purpose of this research work is to design and build a brain mapping based stress identi cation system using single electrode EEG device. Here we have taken EEG recording of 63 students using a single electrode EEG device NeuroSky Mindwave Mobile. A novel feature combination of bandpower ratios of alpha, beta, delta and theta bands are extracted and fed to the classi ers. We evaluate the performance of Support Vector Machine and K-Nearest Neighbour machine learning algorithms. Target class of the training set is assigned using stress score calculated from the response of Perceived Stress Score (PSS-14) questionnaire. We achieve the highest average classi cation accuracy of 74.43% using K-NN algorithm. We observe that there exist a correlation of bandpower ratios of di erent bands obtained from recorded EEG signals from the frontal portion of the brain with the stress level of subjects.en_US
dc.language.isoenen_US
dc.relation.ispartofseriesTD-2794;-
dc.subjectBRAIN MAPPINGen_US
dc.subjectSTRESS IDENTIFICATIONen_US
dc.subjectEEGen_US
dc.subjectBCIen_US
dc.titleBRAIN MAPPING BASED STRESS IDENTIFICATIONen_US
dc.typeThesisen_US
Appears in Collections:M.E./M.Tech. Computer Engineering

Files in This Item:
File Description SizeFormat 
2k15cse14_major_report_brain_mapping_based_stress_identifica.pdf2.61 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.