Distinguishing Hearts: How Machine Learning identifies People based on their Heartbeat

Abstract

Though biometrics are moving into a recent focus, they are actually the oldest form of identification. Humans, and even some animals, recognize each other by their voice, body shape and face. But with the emergence of sensors close to the body combined with the possibilities of Artificial Intelligence (AI), other factors such as the gait and behaviorals are also becoming of increasingly interest.Therefore, this paper illustrates how individuals, supported by Machine Learning (ML) methods, can be distinguished based on their Electrocardiogram (ECG) signals. ECG values recorded with an Microcontroller Unit (MCU) are used and the applicability of three different ML methods -K-Nearest Neighbor (KNN), Support Vector Machine (SVM) and Gaussian Naive Bayes (GNB)- are compared. The results also indicate the potential of ML in terms of applications in (tele)medicine and disease prevention.