Article Per Year (5 Year)
p-Index From 2019 - 2024
0.23
P-Index
This Author published in this journals
All Journal JURNAL MEDIA INFORMATIKA BUDIDARMA
Erlina Agustin
Universitas Muhammadiyah Sidoarjo, Sidoarjo
Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Deteksi Penyakit Epilepsi Melalui Sinyal EEG Menggunakan Metode DWT dan Extreme Gradient Boosting Erlina Agustin; Ade Eviyanti; Nuril Lutvi Azizah
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 7, No 1 (2023): Januari 2023
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v7i1.5412

Abstract

Epilepsy is a disorder of the central nervous system due to excessive patterns of electrical activity in the brain. This disease causes patients to experience repeated seizures in one or all parts of the body. Therefore, epilepsy must be detected early so that the patient immediately gets the right treatment so that the condition does not get worse. This study proposes the detection of epilepsy using the Discrete Wavelet Transform method for feature extraction and Extreme Gradient Boosting for classification. Detection results are classified into two classes, namely seizures and non-seizures. The EEG recording data used came from CHIB MIT Hospital Boston which was obtained online. In the classification process, this study uses four comparisons of the percentage of training data and test data as well as tuning parameters which are processed by Randomized Search Cross Validation. The combination of these methods produces the highest accuracy, namely 85.15% which is produced by the percentage of 80% training data and 20% test data. However, these results experienced a high overfitting of 13.54%. As for the most fit results produced by the research, namely an accuracy value of 81% with a training score of 88.65% and a test score of 81.20% resulting from a percentage of 80% training data and 20% test data.
Co-Authors Ade Eviyanti Nuril Lutvi Azizah