Cardiac Arrhythmia Classification using Ensemble Machine Learning Algorithms with PPG and ECG signals

In this study, machine learning calculations such as Support Vector Machine (SVM), K-Nearest Neighbours (KNN), Logistic Regression, and Random Forest are explored to classify cardiac arrhythmias utilizing Electrocardiogram (ECG) and Photoplethysmogram (PPG) signals. The aim is to create models that not only precisely identify peaks but also analyze cardiac arrhythmias, thereby improving the management of cardiovascular disorders. A thorough literature review was conducted to determine the best approach to treat ECG and PPG signals, leading to the design of calculations to differentiate R-peaks from ECG signals and other waveforms from PPG signals. A comparison study involving four models found that Random Forest and Logistic Regression performed significantly better than the other models.