Septadaya, Aditya (2019) Implementasi Extreme Learning Machine dan Fast Independent Component Analysis untuk Klasifikasi Aritmia Berdasarkan Rekaman Elektrokardiogram. Sarjana thesis, Universitas Brawijaya.
Abstract
Jenis aritmia dapat mengindikasikan letak gangguan dan penyebabnya. Cara mengetahui aritmia adalah dengan menggunakan strip elektrokardiogram (EKG). Machine learning dapat digunakan sebagai pendekatan untuk memudahkan identifikasi aritmia melalui EKG. Extreme Learning Machine (ELM) adalah salah satu single-hidden layer feedforward neural networks (SLFNs) yang dapat digunakan untuk klasifikasi aritmia dalam rangka membantu diagnosa medis. Untuk mengoptimalkan performa ELM, digunakan algoritme Fast Independent Component Analysis (FastICA) untuk preprocessing dan ekstraksi sinyal EKG. Dalam penelitian ini, dilakukan beberapa pengujian parameter untuk mengetahui dampaknya terhadap performa model klasifikasi. Data EKG diperoleh dari database aritmia yang dikelola oleh Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIN). Setiap data merupakan cuplikan EKG selama 3 detik dengan keseluruhan data sebanyak 210 dibagi ke dalam 6 kelas aritmia dan ritme normal. Hasil penelitian menunjukkan bahwa model klasifikasi mampu mencapai performa sempurna dengan akurasi, precision-recall, dan F-1 score sebesar 100% pada tahap pelatihan. Namun model klasifikasi mengalami overfitting pada tahap pengujian dengan rerata matthew correlation coefficient sekitar 0. Overfitting terjadi karena representasi fitur terlalu kompleks dan tidak sebanding dengan jumlah data yang ada. Hal ini mengakibatkan performa yang buruk pada pengujian ELM-FastICA untuk data yang belum dikenali.
English Abstract
This type of arrhythmia can indicate the location of the disorder and its cause. The way to find out the arrhythmia is to use an electrocardiogram (EKG) strip. Machine learning can be used as an approach to facilitate identification of arrhythmias through an EKG. Extreme Learning Machine (ELM) is one of the single-hidden layer feedforward neural networks (SLFNs) that can be used for the classification of arrhythmias in order to assist medical diagnosis. To optimize ELM performance, Fast Independent Component Analysis (FastICA) algorithm is used for preprocessing and extracting ECG signals. In this study, several parameter tests were conducted to determine the impact on the performance of the classification model. ECG data obtained from the arrhythmia database managed by the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIN).Each data is an ECG trailer for 3 seconds with a total of 210 data divided into 6 arrhythmia classes and normal rhythms. The results showed that the classification model was able to achieve perfect performance with accuracy, precision-recall, and F-1 score of 100% at the training stage. However the classification model experiences overfitting at the testing stage with an average matrix correlation coefficient of around 0. Overfitting occurs because the representation of features is too complex and not proportional to the amount of data available. This has resulted in poor performance in the ELM-FastICA test for unrecognized data.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FILKOM/2019/235/051903328 |
| Uncontrolled Keywords: | aritmia, sinyal elektrokardiogram, klasifikasi, extreme learning machine, fast independent component analysis-arhythmia, signal electrocardiogram, classification, extreme learning machine, fast independent component analysis. |
| Subjects: | 000 Computer science, information and general works > 001 Knowledge > 001.012 Classification |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | soegeng sugeng |
| Date Deposited: | 20 Jul 2020 05:27 |
| Last Modified: | 19 Oct 2021 08:47 |
| URI: | http://repository.ub.ac.id/id/eprint/169915 |
Preview |
Text
Aditya Septadaya.pdf Download (61MB) | Preview |
Actions (login required)
 |
View Item |