Securing the Biometric through ECG using Machine Learning Techniques

Abstract

In the current era, biometrics is widely used for maintaining the security. To extract the information from the biomedical signals, biomedical signal processing is needed. One of the significant tools used for the diagnostic is electrocardiogram (ECG). The main reason behind this is the certain uniqueness in the ECG signals of the individual.  In this paper, the focus will be on distinguishing the individual on the basis of ECG signals using feature extraction approaches and the machine learning algorithms. Other than preprocessing approach, the discrete cosine transform is applied to perform the extraction. The classification between the signals of the individuals is carried out using the Support Vector Machine and K-Nearest Neighbor machine learning techniques.  The classification accuracy achieved through SVM is 87% and K-NN has achieved a classification accuracy of 96.6% with k=3. The work has shown how machine learning can be used to classify the ECG signal.