PENERAPAN ALGORITMA BACKPROPAGATION PADA KLASIFIKASI KATA BERBASIS SINYAL EEG
Total View This Week0
Total View This Week0
Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Hermanto, Teguh
Yosrita, Efy
Aziza, Rosida Nur
Yosrita, Efy
Aziza, Rosida Nur
Subject
Teknik Informatika
Datestamp
2022-09-27 02:26:03
Abstract :
Locked In Syndrome (LIS) is a condition in which the patient is conscious but unable to move or communicate verbally due to total paralysis. One way to communicate with LIS sufferers is to communicate through brain signal activity or it is called the Electroencephalogram (EEG). This study uses Fast Fourier Transform (FFT) feature extraction to represent the signal in the discrete time domain and frequency domain of the S01-1Aj subject as research data and produces data with 5 classes, namely empty class, picture class, reading class, heart class, imaginary class. . The extracted data will be used for making machine learning models for artificial neural networks (ANN) and word classification using the backpropagation algorithm which will be visualized using an Arduino paired with an LCD. The results of this study produce a word classification application based on EEG signals. In machine learning testing for exercise data, it produced an accuracy rate of 92.5% and in testing on 5 classes the trial data produced the highest accuracy rate of 47%, precision 47% and recall of 46%.
Locked In Syndrome (LIS) is a condition in which the patient is conscious but unable to move or communicate verbally due to total paralysis. One way to communicate with LIS sufferers is to communicate through brain signal activity or it is called the Electroencephalogram (EEG). This study uses Fast Fourier Transform (FFT) feature extraction to represent the signal in the discrete time domain and frequency domain of the S01-1Aj subject as research data and produces data with 5 classes, namely empty class, picture class, reading class, heart class, imaginary class. . The extracted data will be used for making machine learning models for artificial neural networks (ANN) and word classification using the backpropagation algorithm which will be visualized using an Arduino paired with an LCD. The results of this study produce a word classification application based on EEG signals. In machine learning testing for exercise data, it produced an accuracy rate of 92.5% and in testing on 5 classes the trial data produced the highest accuracy rate of 47%, precision 47% and recall of 46%.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
