Document Details

Document Type : Thesis 
Document Title :
DEVELOPMENT OF EEG-BASED DEEP DEEP LEARNING NN FOR STRESS CLASSIFICATION AND TREATMENT
تطوير شبكة عصبية عميقة التعلم لتصنيف ومعالجة اشارات الإجهاد المضمنة في التخطيط الكهربائي للدماغ
 
Subject : Faculty of Engineering 
Document Language : Arabic 
Abstract : The accurate classification of Stress EEG signals is a big challenge, because of the poor internal signal to noise ratio, for example caused by cardiac signals and movement artefacts. In this work, we proposed to design an EEG- Based Deep Learning Neural Network (DLNN) for Stress Classification. In the first task, we aim to develop a first stage spatial filter intended to remove the dependence on the channel layout, second we propose to employ spatiotemporal convolution in the DLNN to capture both spatial and temporal relationships in the Stress EEG signals. Finally, we will use some techniques such as Optimal Brain Damage to reduce the architecture of the DLNN. As a result, the signal-to-noise (SNR) ratio will be improved, and both the dimensionality of the Stress EEG signal and complexity of DLNN will be reduced. The ultimate aim is to employ the designed DLNN for classification and treatment of different levels of stress. 
Supervisor : Dr. Mohammed Moinuddin 
Thesis Type : Master Thesis 
Publishing Year : 1440 AH
2018 AD 
Added Date : Monday, December 31, 2018 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
عبد العزيز بو طالبBoutalb, Abdelaziz ResearcherMaster 

Files

File NameTypeDescription
 43884.pdf pdf 

Back To Researches Page