Arrhythmia Classification with ECG Signal using Extreme Gradient Boosting (XGBoost) Algorithm
This study aims to fill the research gap by creating an ECG classification model to detect arrhythmia using the XGBoost algorithm, and the results are quite good for each class, but are still considered not better than those models.
Abstract
Heart disease is one of the most dangerous illnesses because it has the potential to take people's lives. One of the causes of heart disease is arrhythmia, an abnormal condition of the heartbeat. To diagnose arrhythmia, analysis of electrocardiographic (ECG) signals can be performed. However, this analysis is very difficult to do conventionally and has the potential for errors, so there is a need for automatic ECG classification to detect arrhythmia. This study aims to fill the research gap by creating an ECG classification model to detect arrhythmia using the XGBoost algorithm. The results are quite good for each class, with accuracies for class N at 98.87%, class SVEB at 99.37%, class VEB at 99.4%, class F at 99.75%, and class Q at 99.99%. However, compared to existing methods in previous research, these results are still considered not better than those models.