Identification of ElectroEncephaloGraph signals using Sampling Technique and K - Nearest Neighbor

ade, efiyanti (2019) Identification of ElectroEncephaloGraph signals using Sampling Technique and K - Nearest Neighbor. Journal of Physics: Conference Series.

Text C. Identification of ElectroEncephaloGraph signals using Sampling Technique and K - Nearest Neighbor.pdf Download (547kB)

Abstract

Electroencephalograph is a device that can help humans observe and analyze the results of electrical waves produced by neurons in the brain. The results of reading the tool are called Electroencephalogram (EEG), besides being able to help diagnose physician for medical therapy, it is also developed for the Brain-Computer Interfacing (BCI) application. BCI is a method that allows humans to be able to control an external system without direct contact with the system. Research on communication between humans to control external equipment has been widely investigated, including research on brain activity to control a cursor on a computer screen. This study focuses on feature extraction for ElectroEncephaloGraph (EEG) signals using the sampling technique. K - Nearest Neighbor is used as a classification of EEG signals to determine whether the cursor moves up or down. The data used are EEG data originating from the 2003 BCI competition (BCI 2003 Competition). Decision making is done to classify the cursor movement up and down the cursor movement. The research data uses 250 EEG signal file training data and 50 from EEG signal file testing data, so that the whole becomes 300 EEG signal data files. The best results with K = 3 values obtained for the classification of EEG signals using K-NN are 76% of the signal data tested

Item Type:	Other
Subjects:	T Technology > T Technology (General)
Divisions:	Faculty of Engineering > School of Computer Engineering
Depositing User:	mr hindarto hindarto
Date Deposited:	26 Jan 2021 07:15
Last Modified:	26 Jan 2021 07:15
URI:	http://eprints.umsida.ac.id/id/eprint/8220

Actions (login required)

View Item