Klasifikasi Aritmia Dari Hasil Elektrokardiogram Menggunakan Support Vector Machine Dengan Seleksi Fitur Menggunakan Algoritma Genetika

Reiza Adi Cahya (Fakultas Ilmu Komputer, Universitas Brawijaya)
Candra Dewi (Fakultas Ilmu Komputer, Universitas Brawijaya)
Bayu Rahayudi (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 2 No 3 (2018): Maret 2018

Electrocardiogram (ECG) can be used to recognize abnormal heart beats or arrhythmia. Automatic arrhythmia recognition can be achieved through the use of machine learning techniques. However, ECG generates raw numerical data with large amount of features that can reduce the quality of automatic recognition. Genetic algorithm (GA) can be utilized to perform a feature selection, reducing the amount of features. Data with reduced features then will be used to train a support vector machine (SVM) classifier. ECG data from the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database is used as training and testing data. Each data is a six-second ECG recording, and is classified into normal heartbeat and 3 different kind of arrhythmias. Result shows that GA-SVM yielded average accuracy of 82.5% with 120 training data and 20 test data, and reduced the amount of feature from 2160 original features to an average of 406 reduced features.