Document Details

Document Type : Thesis 
Document Title :
A DE-SVM FEATURE SELECTION MODEL BASED ON HIGH PERFORMANCE COMPUTING (HPC) TECHNIQUES FOR P300 BASED BRAIN COMPUTER INTERFACE (BCI) DATA
نموذج اختيار الميزة DE-SVM القائم على تقنيات الحوسبة العالية الأداء (HPC) لبيانات واجهة الدماغ الحاسوبية (BCI) القائمة على P300
 
Subject : Faculty of Computing and Information Technology 
Document Language : Arabic 
Abstract : The applications of High Performance Computing (HPC) have been a field of interest in many different disciplines. HPC proved notable performance enhancements. This research is considering the Brain Computer Interface (BCI) dataset, precisely the P300 based system. BCI is a system provides a direct communication control channel between the brain and the external world, but its data processing is exceedingly time consuming. That system consists of many components where Feature Selection is a primary key of its performance. Search algorithms and classifiers form the feature selection model. Hence, they are the concern of this research where Genetic Algorithm (GA) and Differential Evolution (DE) implemented as search algorithms. These Evolutionary Algorithms (EA) estimate an optimal solution saving the enormous amount of time consumed by a brute force search. Linear Discriminant Analysis (LDA) and Support Vector Machine (SVM) are the classifiers used. Thus, there are four models: GA-LDA, DE-LDA, GA-SVM, and DE-SVM. HPC techniques implemented since the computational power was one of the main obstacle beside the problem's size causing an extensive processing time. DE-SVM model proved to be the best where it saves 98.8% of the original time consumed while using ordinary computing facilities. It also maintains an accuracy rate of almost 80% selecting 42% of the original features only 
Supervisor : Dr. Mohamed Dahab 
Thesis Type : Master Thesis 
Publishing Year : 1439 AH
2018 AD 
Co-Supervisor : Dr. Mahmoud Kamel 
Added Date : Monday, January 22, 2018 

Researchers

Researcher Name (Arabic)Researcher Name (English)Researcher TypeDr GradeEmail
سحر عائض الوادعيAlwadei, Sahar AyedhResearcherMaster 

Files

File NameTypeDescription
 43023.pdf pdf 

Back To Researches Page