Indonesia Conference Directory


<< Back

Features Selection and Classification of Motor Imagery using Particel Swarm Optimization and SVM Algorithm
Yabes Dwi Nugroho(a*), Prisma Megantoro (a), Iswanto (b*)

Universitas Gadjah Mada
Departement of Electrical Engineering and Information Technology
yabes.mti15[at]mail.ugm.ac.id
prisma.megantoro[at]mail.ugm.ac.id

Universitas Muhammadiyah Yogyakarta
Electrical Engineering
iswanto_te[at]umy.ac.id


Abstract

The nervous system in the body has information to move body parts. This information will be classified to find out what signal information is in the human brain. One developing application is Electroenchephalography (EEG), which is an instrument for capturing these brain signals. Signal patterns can be known from the stimulation a person receives from movements that result from brain activity. In this study, the authors built a system that would classify imagery motor patterns of brain activity through the EEG produced. CSP is used as a feature selection to retrieve patterns from imagery motor signals. Meanwhile, at the classification stage use PSO and SVM. PSO is used because it is more efficient because it requires less computing. Based on the results of calculations carried out it produces a significant accuracy of 93%.

Keywords: Feature Selection, Classification, EEG, Motor Imagery, CSP, PSO, SVM

Topic: International Symposium of Engineering, Technology, and Health Sciences

Link: https://ifory.id/abstract/4RVYkGrxvyLh

Conference: The 3rd International Conference on Sustainability and Innovation (ICoSI 2019)

Plain Format | Corresponding Author (Yabes Dwi Nugroho H)

Featured Events

<< Swipe >>
<< Swipe >>

Embed Logo

If your conference is listed in our system, please put our logo somewhere in your website. Simply copy-paste the HTML code below to your website (ask your web admin):

<a target="_blank" href="https://ifory.id"><img src="https://ifory.id/ifory.png" title="Ifory - Indonesia Conference Directory" width="150" height="" border="0"></a>

Site Stats