Harahap, Naufal Rafi Akbar (2022) Penerapan Satu Dimensi Convolutional Neural Network Dan Deep Neural Network Untuk Klasifikasi Penyakit Epilepsi. Other thesis, Institut Teknologi Sepuluh Nopember.
![]() |
Text
05111840000157-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (2MB) | Request a copy |
Abstract
Epilepsi adalah penyakit otak kronis umum yang disebabkan oleh aktivitas saraf abnormal dan terjadinya kejang mendadak atau sementara. Electroencephalogram (EEG) adalah teknik non-invasif yang biasa digunakan untuk mengidentifikasi aktivitas otak epilepsi. Deteksi visual EEG bersifat subjektif dan memakan waktu untuk ahli saraf. Oleh karena itu, penulis mengusulkan deteksi penyakit epilepsi otomatis. Fitur sinyal EEG diekstraksi menggunakan Discrete Fourier Transform (DFT) dan Discrete Wavelet Transform (DWT) yang kemudian fitur tersebut akan diseleksi menggunakan XGBoost untuk menghilangkan fitur yang tidak relevan, dan pengulangan. Mesin klasifikasi menggunakan satu dimensi Convolutional Neural Network (CNN) dan Deep Neural Network (DNN). Hasil penelitian membuktikan dengan DWT berhasil diimplementasikan untuk melakukan ekstraksi fitur, akurasi dari tiap kelas bergantung dengan tipe famili wavelet dan level dekomposisi. Penambahan lapisan DNN dari hasil klasifikasi CNN berhasil meningkatkan akurasi dari tiap jenis kelas klasifikasi, membuktikan penambahan lapisan DNN berhasil melakuan filterisasi dengan baik.
=================================================================================================================================
Epilepsy is a common chronic brain disease caused by abnormal nerve activity and the occurrence of sudden or transient seizures. Electroencephalogram (EEG) is a non-invasive technique commonly used to identify epileptic brain activity. EEG visual detection is subjective and time consuming for the neurologist. Therefore, the authors propose automatic detection of epilepsy. EEG signal features are extracted using Discrete Fourier Transform (DFT) and Discrete Wavelet Transform (DWT) which will then be selected using XGBoost to eliminate irrelevant features and repetition. The classification uses one-dimensional Convolutional Neural Network (CNN) and Deep Neural Network (DNN). The results of this study prove that DWT has been successfully implemented to perform feature extraction, the accuracy of each class depends on the type of wavelet family and the level of decomposition. The addition of the DNN layer from the results of the CNN classification has succeeded in increasing the accuracy of each type of classification class, proving that the addition of the DNN layer has successfully filtered well.
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | Epilepsi, sinyal EEG, DFT, DWT, satu-dimensi CNN, DNN, Epilepsy, EEG Signals, One-dimensional |
Subjects: | T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing |
Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis |
Depositing User: | Mr. Marsudiyana - |
Date Deposited: | 13 Oct 2025 05:29 |
Last Modified: | 13 Oct 2025 05:29 |
URI: | http://repository.its.ac.id/id/eprint/128576 |
Actions (login required)
![]() |
View Item |