Electrocardiogram based Gender Classification

ECG signal having distinguishable features is used for the classification of gender. This paper presents time and frequency domain features for the classification of gender. A system is designed which can accurately distinguish between genders using an ECG signal. This study gives us a complete framework of gender classification based on ECG signals. The raw ECG signals taken from Physio Bank are pre-processed using empirical mode decomposition (EMD) are used. Kernel PCA is used for reducing the features from 42 to 7. Fine tree classifier has been used to distinguish between different classes of genders. In this particular study, two classes i.e. males and females have been used and obtained a maximum accuracy of 95.2%. This result shows that it is a competent system that can differentiate between different genders.